A tornado cooperative observation method based on a detection device

Abstract

The application relates to the technical field of meteorological detection, in particular to a tornado cooperative observation method based on a detection device. The method comprises the following steps: performing real-time monitoring on regional micro-pressure and meteorological echoes of a suspicious target area through a micro-pressure gauge and an S-band new-generation weather radar, and performing strong convective occurrence matching prediction to obtain a strong convective occurrence prediction target area; performing solar radio calibration processing on corresponding radar detection equipment and triggering a tornado detection strategy cooperative instruction to perform strong convective target three-dimensional cooperative observation and target strong correlation area fusion division, so as to generate a strong convective target and strong correlation area radar field; performing strong convective feature analysis on the strong convective target and strong correlation area radar field, so as to generate a tornado and strong convective feature data set; acquiring a preset tornado feature library and performing tornado recognition analysis, so as to generate a tornado occurrence prone grid area. The application can improve the monitoring accuracy and real-time performance of the tornado detection process.

Description

Technical Field

[0001] The present invention relates to the field of meteorological detection technology, and in particular to a tornado collaborative observation method based on detection equipment. Background Art

[0002] Currently, tornado observation methods primarily rely on equipment such as ground-based radar, weather radar, anemometers, and weather rovers. Ground-based radar provides real-time data on tornadoes by detecting parameters such as wind speed, direction, and rainfall. Weather radar monitors tornado formation and path from high altitudes. Anemometers and weather rovers can monitor meteorological changes near the ground in real time. Furthermore, current research is shifting toward multi-device collaborative observation models. By combining multiple detection devices, multi-angle, multi-level, and real-time monitoring of the entire tornado process can be achieved, thereby improving tornado prediction and early warning capabilities. However, existing multi-device collaborative observation methods primarily focus on data fusion and information sharing between different devices. Furthermore, the collaborative working methods and data synchronization mechanisms between devices remain inadequate, making it difficult to achieve efficient and seamless collaboration between detection devices in practical applications. This hinders the timely provision of accurate tornado observation data, resulting in insufficient timeliness and accuracy in observation results. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a tornado collaborative observation method based on detection equipment to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a tornado collaborative observation method based on detection equipment includes the following steps:

[0005] Step S1: Using a micromanometer and a new generation S-band weather radar, the suspected target area is monitored for regional micropressure and meteorological echoes in real time to obtain micropressure change data and meteorological echo data in the suspected area; based on the micropressure change data and meteorological echo data in the suspected area, a strong convection occurrence matching prediction is performed on the suspected target area to obtain a predicted target area for strong convection occurrence;

[0006] Step S2: performing solar radio calibration on the corresponding radar detection equipment in the tornado detection network to generate a tornado detection calibration optimization network; triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspected target in the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspected target in the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target;

[0007] Step S3: performing target strong correlation region fusion division on the corresponding strong convective suspicious targets in the three-dimensional radar field of the strong convective target to generate a radar field of strong convective target and strong correlation region; performing isotropic convection feature analysis on the radar field of strong convective target and strong correlation region using a spatiotemporal adaptive reflectivity fusion model to generate a dataset of strong convective features such as tornadoes;

[0008] Step S4: Obtain a preset tornado feature library, and perform tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado feature library and the tornado and other strong convective feature data set to generate a tornado-prone grid point area.

[0009] Furthermore, step S1 includes the following steps:

[0010] Step S11: Using a micromanometer, real-time monitoring of regional micropressure in the suspicious target area is performed according to monitoring conditions corresponding to a long period of 100-250 minutes and an amplitude fluctuation greater than or equal to 60 Pa, or a short period of 20-99 minutes and an amplitude fluctuation greater than or equal to 20 Pa, to obtain micropressure change data of the suspicious area;

[0011] Step S12: Using a new generation S-band weather radar, the suspicious target area is monitored for regional meteorological echoes in real time to obtain meteorological echo data of the suspicious area;

[0012] Step S13: performing spatiotemporal joint filtering on the micro-pressure change data of the suspicious area and the meteorological echo data of the suspicious area to obtain the micro-pressure filtered data and meteorological filtered data of the suspicious area corresponding to each spatiotemporal position point;

[0013] Step S14: performing a matching quantitative calculation on the suspicious target area using a regional strong convection feature matching calculation formula based on the corresponding suspicious area micro-pressure filter data and the suspicious area meteorological filter data at each time and space position point to obtain a strong convection feature matching degree of the suspicious area;

[0014] Step S15: Based on the strong convection feature matching degree of the suspicious area, a strong convection occurrence matching prediction is performed on the corresponding strong convection suspicious target in the suspicious target area. If the strong convection feature matching degree of the suspicious area is greater than or equal to 70%, the corresponding strong convection suspicious target in the suspicious target area is displayed, otherwise it is not displayed, so as to obtain the strong convection occurrence prediction target area.

[0015] Furthermore, step S12 includes the following steps:

[0016] Step S121: transmitting corresponding S-band electromagnetic waves by a new generation S-band weather radar to perform regional meteorological echo scanning on the meteorological components behind the atmosphere in the suspicious target area to generate echo signals of the meteorological components in the suspicious area;

[0017] Step S122: analyzing the signal strength and reflectivity of the meteorological component echo signal in the suspicious area to obtain meteorological echo signal strength data and meteorological echo signal reflectivity data in the suspicious area;

[0018] Step S123: performing a signal echo spatiotemporal distribution analysis of the meteorological components behind the atmosphere in the suspicious target area based on the suspicious area meteorological echo signal strength data and the suspicious area meteorological echo signal reflectivity data to generate a suspicious area meteorological signal echo spatiotemporal distribution map;

[0019] Step S124: performing a meteorological spatiotemporal distribution field simulation analysis on the suspicious target area according to the spatiotemporal distribution map of meteorological signal echoes in the suspicious area to generate a spatiotemporal distribution field of meteorological echoes in the suspicious area;

[0020] Step S125: Real-time monitoring of the spatiotemporal distribution of meteorological echoes in the suspicious area is performed to obtain meteorological echo data in the suspicious area, wherein the meteorological echo data in the suspicious area includes corresponding meteorological conditions and cloud density changes in the suspicious target area.

[0021] Furthermore, the regional severe convection feature matching calculation formula in step S14 is specifically:

[0022] ;

[0023] Where, For suspicious target areas at time and location points The matching degree of strong convection characteristics in the corresponding suspicious area is: is the spatial domain of the suspicious target area, is the start time of the time range, is the end time of the time range, For suspicious target areas at time and location points The corresponding micro-pressure parameters of the suspicious area are: For suspicious target areas at time and location points The corresponding meteorological parameters of the suspicious area are as follows: is the characteristic parameter field of severe convection in the suspicious target area, is the weight coefficient of the strong convection characteristics, For suspicious target areas at time and location points The corresponding micro-pressure field gradient is: For suspicious target areas at time and location points The corresponding meteorological field gradient is is the gradient adjustment coefficient of the strong convection characteristic field, is an exponential function, is the reference position of the center of the suspicious target area, is the spatial position attenuation coefficient, is the correction coefficient for the matching degree of severe convective characteristics in the suspicious area.

[0024] Furthermore, step S2 includes the following steps:

[0025] Step S21: obtaining the solar radio radiation intensity at the corresponding signal wavelength of the corresponding radar detection equipment in the tornado detection network under the current solar radiation;

[0026] Step S22: Calculate the received signal power of the corresponding radar detection equipment in the tornado detection network based on the solar radio radiation intensity at the corresponding signal wavelength under the current solar radiation to obtain the theoretical value of the radar detection received signal power. The calculation formula of the theoretical value of the radar detection received signal power is specifically:

[0027] ;

[0028] in, is the theoretical value of radar detection received signal power, is the radar receiving frequency bandwidth, is the radar antenna gain, is the radar signal wavelength, The wavelength of the radar signal The intensity of solar radio radiation at The radar receiving signal frequency;

[0029] Step S23: obtaining a measured value of the radar detection received signal power, and performing a power difference calculation based on the measured value of the radar detection received signal power and a theoretical value of the radar detection received signal power to obtain a radar detection received signal power difference;

[0030] Step S24: performing solar radio calibration processing on the corresponding radar detection equipment in the tornado detection network based on the difference in radar detection received signal power to generate a tornado detection calibration optimization network;

[0031] Step S25: triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspicious target within the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspicious target within the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target.

[0032] Furthermore, step S25 includes the following steps:

[0033] Step S251: triggering the built-in policy control server to respond and send a tornado detection policy coordination instruction based on the existence of a suspected strong convection target corresponding to the strong convection occurrence prediction target area;

[0034] Step S252: applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network and generating a tornado detection coordination scanning command in response;

[0035] Step S253: applying the tornado detection collaborative scanning command to the corresponding search radar in the tornado detection calibration optimization network to perform target area detection scanning on the corresponding strong convection suspected targets within the strong convection occurrence prediction target area, so as to generate the target scanning signal echo intensity corresponding to each strong convection occurrence target area;

[0036] Step S254: applying the tornado detection collaborative scanning command to the corresponding optical camera in the tornado detection calibration optimization network to perform target synchronous acquisition scanning on the corresponding strong convection suspected targets in the strong convection prediction target area to generate the target area size and target area position corresponding to each strong convection target area; calculating the scanning azimuth angle range based on the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convection target area to generate the strong convection target area scanning azimuth angle distribution range;

[0037] Step S255: Apply the tornado detection collaborative scanning command to the corresponding ultra-fine radar in the tornado detection calibration optimization network and perform fixed-point collaborative fan scanning tracking and three-dimensional analysis on the corresponding strong convection suspected targets in the strong convection prediction target area based on the scanning azimuth distribution range of the strong convection target area to generate a three-dimensional radar field of the strong convection target.

[0038] Furthermore, the calculation of the scanning azimuth angle range according to the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convection target area in step S254 includes the following steps:

[0039] Performing spatial coordinate system conversion according to the target area position corresponding to each severe convection target area to generate a severe convection target area position spatial coordinate system;

[0040] Based on the spatial coordinate system of the strong convection target area, the target area size corresponding to each strong convection target area is measured and calculated to obtain the regional scale characteristic index corresponding to each strong convection target area, where the regional scale characteristic index includes the regional major axis, regional minor axis and regional total area size;

[0041] Based on the spatial coordinate system of the strong convection target area, the scanning azimuth angle range of the target scanning signal echo intensity and regional scale characteristic indicators corresponding to each strong convection target area are calculated to generate the scanning azimuth angle distribution range of the strong convection target area.

[0042] Furthermore, step S3 includes the following steps:

[0043] Step S31: obtaining a severe convective weather activity characteristic data matrix and an environmental condition distribution climate characteristic data matrix corresponding to the severe convective suspicious target in the severe convective target three-dimensional radar field;

[0044] Step S32: Based on the severe convective weather activity characteristic data matrix and the environmental condition distribution climate characteristic data matrix, a canonical correlation analysis expression is used to perform a canonical correlation calculation on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field to obtain a canonical correlation coefficient between the strong convective weather activity corresponding to the strong convective suspicious targets and the environmental conditions;

[0045] Among them, the canonical correlation analysis expression is specifically:

[0046] ;

[0047] Where, is the canonical correlation coefficient, is the coefficient vector corresponding to the typical variables in the characteristic data matrix of severe convective weather activities, is the coefficient vector corresponding to the typical variables in the climate characteristic data matrix of environmental condition distribution, is the vector transpose symbol, is the characteristic data matrix of severe convective weather activities, is the climate characteristic data matrix of environmental conditions distribution, is the expected value, is the covariance matrix between severe convective weather activities and environmental conditions, is the autocovariance matrix of severe convective weather activity data, is the autocovariance matrix of environmental condition data;

[0048] Step S33: performing target equal-correlation classification on the corresponding strong convective suspicious targets in the three-dimensional radar field based on the typical correlation coefficient between the strong convective weather activities corresponding to the strong convective suspicious targets and the environmental conditions, so as to obtain a set of strong convective target equal-correlation;

[0049] Step S34: performing target strong correlation region fusion division on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field based on the strong convective target equal strong correlation set to generate a strong convective target equal strong correlation region radar field;

[0050] Step S35: Gridding the radar field of the strong convective target and other strong correlation areas using the spatiotemporal adaptive reflectivity fusion model to generate a gridded strong convective grid field of the strong correlation areas; performing strong convective feature analysis on the gridded strong convective grid field of the strong correlation areas to generate a tornado and other strong convective feature dataset, wherein the tornado and other strong convective feature dataset includes horizontal polarization radar reflectivity. , horizontal and vertical dual polarization differential reflectivity , radar echo correlation coefficient , horizontal and vertical dual polarization rotation angle And the horizontal and vertical dual polarization differential phase standard deviation ,in is the differential phase measured by the radar under horizontal and vertical dual polarization.

[0051] Furthermore, the expression corresponding to the spatiotemporal adaptive reflectivity fusion model in step S35 is specifically:

[0052] ;

[0053] Where, Output value of the spatiotemporal adaptive reflectivity fusion model, indicating the center position of the grid and time The gridded strong convective reflectivity value at is the maximum number of grid points, is the item index of the horizontal grid point, is the item index of the vertical grid point, Horizontal grid With vertical grid The weight coefficient matrix between For the The horizontal position parameters corresponding to the horizontal grid points, For the The vertical position parameters corresponding to the vertical grid points, At the grid point and time The original radar observed reflectivity value at At the grid point and the previous moment The original radar observed reflectivity value at At the grid point and the previous moment Gridded strong convective reflectivity values ​​at .

[0054] Furthermore, step S4 includes the following steps:

[0055] Step S41: obtaining a preset tornado feature library;

[0056] Step S42: Construct a tornado and other strong convection identification model based on the tornado feature library and the tornado and other strong convection feature data set, and calculate the dual-polarization tornado vortex characteristic value based on the tornado and other strong convection identification model and tornado debris characteristic values ;

[0057] Step S43: The dual polarized tornado vortex characteristic value and tornado debris characteristic values Substitute the tornado characteristic membership function corresponding to the severe convection identification model such as tornado to perform membership metric calculation to obtain the severe convection target tornado characteristic membership degree;

[0058] Step S44: performing tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado characteristic membership of the strong convective target to generate a grid point area prone to tornado occurrence.

[0059] Beneficial effects of the present invention:

[0060] The tornado collaborative observation method based on detection equipment proposed by the present invention has the beneficial effect of being able to capture tiny air pressure changes in the area with high precision by using a micromanometer to monitor micropressure fluctuations in a suspicious target area, compared with the existing technology. These changes are often closely related to the evolution of atmospheric physical processes and weather systems, especially the occurrence of severe convective weather phenomena. The micropressure changes monitored by the micromanometer can not only reflect pressure fluctuations in the atmosphere, but also reveal potential meteorological anomalies and convective activities, thereby improving the accuracy of severe convective weather warnings. In addition, by using a new generation of S-band weather radar to conduct real-time monitoring of regional meteorological echoes in suspicious target areas, it is possible to effectively capture severe convective activities, precipitation systems, and unstable layers in the atmosphere in the area. The advantage of S-band radar is its strong penetrating power. It can identify information such as the structure, movement speed, and direction of thunderstorm clouds through the intensity and pattern of echo signals, thereby providing data support for accurate warnings of severe convection. At the same time, by matching the occurrence of severe convection in the suspected target area with the micro-pressure change data and the meteorological echo data in the suspected area, regional severe convective weather warnings can be achieved. The core of this step is to set a threshold (for example, a matching degree greater than or equal to 70%). When this threshold is exceeded, it can be regarded as a signal of impending severe convective weather. The matching degree of severe convection characteristics, as a basis for judgment, can reflect the degree of correlation between regional meteorological conditions and the occurrence of severe convective weather, helping meteorological departments optimize the coverage of the warning area and reduce the losses caused by severe convective weather. Secondly, by obtaining the solar radio radiation intensity at the corresponding signal wavelength under the current solar radiation of the corresponding radar detection equipment in the tornado detection network, the corresponding radar detection equipment is calibrated by solar radio. The radar system calibration is based on the precise measurement of signal power differences. Solar radio calibration is mainly used to eliminate the influence of solar radiation on the radar signal. By analyzing the difference in received signal power, correction can be made for different solar radiation conditions, thereby optimizing the performance of the radar detection equipment, enabling it to maintain stable detection accuracy and more accurately identify and predict severe convective weather phenomena such as tornadoes.It also triggers the tornado detection strategy coordination instruction according to the corresponding strong convection suspicious targets in the strong convection prediction target area, and applies the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of strong convection targets corresponding to the strong convection suspicious targets in the strong convection prediction target area. The core of this step is to trigger the corresponding detection strategy based on the suspicious targets in the strong convection prediction area, and further refine the detection of strong convection phenomena such as tornadoes through three-dimensional coordinated observation. By reasonably setting the coordination instruction, multiple radar devices can work together to conduct more comprehensive and accurate monitoring of the target area. This not only improves the spatial resolution of the detection, but also enhances the temporal resolution, so that the radar detection equipment can conduct all-round and three-dimensional observations in the target area, greatly improving the response capability of the radar system under complex weather conditions, and providing basic data guarantee for the subsequent tornado observation process, thereby realizing efficient and seamless collaborative work between detection equipment. Then, by fusing and dividing the target strongly correlated areas of the strong convective suspicious targets corresponding to the three-dimensional radar field of the strong convective target, the key to this process is that by fusing and dividing the target strongly correlated areas, different types of strong convective weather activities can be effectively integrated spatially, and strong convective activity areas of equal strength in the region can be identified. This division not only helps to improve the spatial resolution of radar data, but also further enhances the spatial recognition capability of strong convective weather. It can accurately locate the occurrence area of ​​equal strength convective weather, and combine historical data with typical characteristics to conduct comparative analysis between regions, thereby providing a more accurate spatial reference for subsequent warning and analysis. This step effectively optimizes the data processing process, makes the regional division of strong convective weather more scientific, and can provide accurate observation data of equal strength tornadoes in a timely manner. The radar field of strongly correlated areas such as strong convective targets is gridded by using a spatiotemporal adaptive reflectivity fusion model, and strong convective characteristics are analyzed. The key to this process is that it can accurately capture the detailed characteristics of strong convective weather through gridding processing and generate a radar field model with spatiotemporal adaptive capabilities. Gridding processing helps to make detailed spatial divisions of strong convective areas, so that radar reflectivity data can be analyzed at a higher spatial resolution, providing accurate input for subsequent extraction of strong convective weather characteristics. By analyzing the strong convective characteristics, especially through the calculation of multiple indicators including horizontal polarization reflectivity, dual-polarization differential reflectivity, and echo correlation coefficient, the specific characteristics of strong convective weather can be better revealed. This process can not only provide a more comprehensive tornado meteorological characteristic data set, but also provide a scientific basis for subsequent tornado identification warnings and risk assessments, thereby improving the timeliness and accuracy of tornado observation results.Finally, by obtaining the preset tornado feature library, which contains various basic data and historical records about tornado occurrence and characteristic indicators of different types of tornadoes, including wind speed, rotation intensity, pressure gradient, radar echo characteristics, dual-polarization radar signal characteristics, etc., and by performing tornado identification analysis on the radar fields of strongly correlated areas such as strong convective targets based on the tornado feature library and strong convective feature data sets such as tornadoes, potential tornado occurrence areas can be effectively identified. Especially when strong convective weather occurs, the performance of the radar field can often reveal signs of tornado formation in advance, which can help determine which radar field areas have a higher probability of tornado occurrence, so as to focus on monitoring these areas. This analysis method can not only improve the recognition rate of tornadoes, but also improve the response speed and accuracy of the tornado observation and warning process, thereby enhancing the ability to respond to extreme weather. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0062] Figure 1 Schematic diagram of the steps of the tornado collaborative observation method based on detection equipment of the present invention;

[0063] Figure 2 for Figure 1 Detailed step flow chart of step S1 in FIG. DETAILED DESCRIPTION

[0064] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0065] To achieve this, please refer to Figures 1 to 2 The present invention provides a tornado collaborative observation method based on detection equipment. In the embodiment of the present invention, please refer to Figure 1 FIG. 1 is a flow chart showing the steps of a tornado collaborative observation method based on detection equipment according to the present invention. In this example, the tornado collaborative observation method based on detection equipment includes the following steps:

[0066] Step S1: Using a micromanometer and a new generation S-band weather radar, the suspected target area is monitored for regional micropressure and meteorological echoes in real time to obtain micropressure change data and meteorological echo data in the suspected area; based on the micropressure change data and meteorological echo data in the suspected area, a strong convection occurrence matching prediction is performed on the suspected target area to obtain a predicted target area for strong convection occurrence;

[0067] In an embodiment of the present invention, a micromanometer is used to monitor the micropressure of a suspected target area in real time. The micromanometer should be set at different locations in the target area and configured to monitor long-period amplitude fluctuations with a period of 100 to 250 minutes, and the amplitude of the fluctuation should be no less than 60Pa. At the same time, the micromanometer should also be able to monitor short-period amplitude fluctuations with a period of 20 to 99 minutes, and the amplitude should be greater than or equal to 20Pa. During operation, the micromanometer records the micro-pressure data in the target area through periodic sampling, thereby obtaining micro-pressure change data in the suspected area. In addition, a new generation of S-band weather radar is used to monitor the meteorological echoes in the suspected target area in real time. The S-band weather radar has high detection accuracy and penetration ability, and can effectively monitor the meteorological echoes in the target area. The radar obtains meteorological change data of the target area in real time by emitting electromagnetic waves and receiving echo signals reflected from meteorological targets, thereby obtaining meteorological echo data of the suspected area. At the same time, by combining the micro-pressure change data and meteorological change data obtained from previous real-time monitoring, the spatial domain of the suspicious target area, and the relevant parameters of the suspicious target area, a suitable feature matching calculation formula is constructed to perform matching quantitative calculations on the suspicious target area, so as to fully consider the relationship between micro-pressure changes and meteorological echo characteristics, as well as the indicative relationship between factors such as reflection intensity and gradient in the echo and strong convective activity. The strong convective feature matching degree at each spatiotemporal position point is quantitatively calculated. This matching degree reflects the probability of whether strong convective weather exists in the suspicious area. Then, the occurrence of the corresponding strong convective suspicious target in the suspicious target area is predicted based on the strong convective feature matching degree obtained previously. When the strong convective feature matching degree in the suspicious area is greater than or equal to 70%, the strong convective suspicious target in the area is automatically displayed, prompting relevant meteorological monitoring personnel or early warning systems to pay attention to the corresponding strong convective weather risk. If the matching degree is lower than the threshold, the strong convective target will not be displayed, indicating that strong convective weather will not occur in the area in the short term. Finally, the target area for strong convection occurrence prediction is obtained.

[0068] Step S2: performing solar radio calibration on the corresponding radar detection equipment in the tornado detection network to generate a tornado detection calibration optimization network; triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspected target in the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspected target in the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target;

[0069] In an embodiment of the present invention, by accurately measuring the impact of solar radiation on radar signals, real-time data of solar radiation intensity is first obtained. During specific implementation, it is necessary to rely on solar radio radiation models and related instruments to measure the solar radiation intensity, and by combining the radar receiving frequency bandwidth, radar antenna gain, radar signal wavelength, solar radio radiation intensity at the corresponding signal wavelength under current solar radiation, radar receiving signal frequency and related parameters, a calculation formula is formed to calculate the receiving signal power of the corresponding radar detection equipment in the tornado detection network to quantify the theoretical power value of the radar receiving signal, and also by obtaining the actual receiving signal power value. These measured values ​​reflect the signal power received by the radar equipment in actual work. After obtaining the measured values, they are compared with the theoretical signal power previously calculated by solar radio to calculate the power difference. ,in is the power difference, is the measured value of signal power, The theoretical value of signal power is taken as the theoretical value, and the corresponding radar equipment in the tornado detection network is calibrated with solar radio by using the power difference calculated previously. In the specific implementation, the deviation value of each device is first calculated according to the power difference of each radar device. This deviation value can reflect the impact of solar radio radiation on different radar devices. Based on these deviation values, the radar equipment is optimized by adjusting parameters such as the radar's receiving sensitivity and gain. This process can be carried out through an automatic calibration system, which includes automated calibration software and hardware interfaces, and can fine-tune the radar equipment in real time so that the radar's received signal power is more in line with the theoretical predicted value. By calibrating all radar equipment, an optimized detection network that can more accurately respond to solar radiation interference is formed, thereby generating a tornado detection calibration optimization network. Then, based on the real-time weather data and radar observation results in the severe convection prediction area, the tornado detection strategy is triggered. Through the meteorological warning system and the severe convection prediction model, the potential severe convection suspected target area is first identified, and this information is used as input to trigger the corresponding tornado detection strategy coordination instructions. Next, the coordinated detection instructions are started according to the optimized tornado detection calibration optimization network, and multiple radar devices are directed to conduct coordinated observations based on the position, speed and direction of the target area. This process transmits coordinated instructions through a real-time data link, allowing multiple radar devices to work synchronously, thus forming a three-dimensional radar field covering the target area. The three-dimensional radar field can provide multi-dimensional monitoring of severe convection targets, including information such as the target's height, speed and spatial distribution, and finally observes and generates a three-dimensional radar field of severe convection targets.

[0070] Step S3: performing target strong correlation region fusion division on the corresponding strong convective suspicious targets in the three-dimensional radar field of the strong convective target to generate a radar field of strong convective target and strong correlation region; performing isotropic convection feature analysis on the radar field of strong convective target and strong correlation region using a spatiotemporal adaptive reflectivity fusion model to generate a dataset of strong convective features such as tornadoes;

[0071] In an embodiment of the present invention, relevant data of strong convective suspicious targets are obtained through the three-dimensional radar field of strong convective targets obtained through previous observations. These radar data are converted into a strong convective weather activity characteristic data matrix, which reflects the intensity, distribution characteristics and change pattern of the radar echo signal. At the same time, the environmental condition distribution climate characteristic data matrix is ​​also collected, which contains multi-dimensional environmental data such as temperature, humidity, air pressure, wind speed, wind direction, etc., and the correlation between the strong convective weather activity characteristics and the environmental condition characteristics is calculated by the canonical correlation analysis (CCA) method. The purpose is to find the canonical correlation coefficient between the two groups of variables, and to classify the strong convective suspicious targets corresponding to the three-dimensional radar field of the strong convective target based on the canonical correlation coefficient obtained by previous quantitative calculation. First, the strong convective suspicious targets are clustered and analyzed using the canonical correlation coefficient obtained previously calculated, so as to classify the targets into different correlation categories by setting a certain correlation threshold. Targets with typical correlation coefficients higher than the threshold are classified into the same strong correlation class, while targets with lower correlation are classified into weak correlation classes, so as to automatically classify strong convective suspicious targets in the radar field into strong correlation classes. Each type of target represents an area with similar weather activity and environmental conditions in the radar field, and realizes the regional division of strong convective targets corresponding to the strong convective suspicious targets in the three-dimensional radar field of strong convective targets, so as to realize the spatial distribution analysis of targets in each strong correlation class, aggregate targets with the same strong correlation into one area, and merge adjacent strong correlation targets into larger areas through spatial clustering technology. In this process, factors such as the target's geographical location, the spatial characteristics of the radar echo, and the environmental conditions will be used as input to ensure that the generated strong convective target and other strong correlation area radar field can accurately reflect the distribution of meteorological activities in different areas, thereby generating a strong convective target and other strong correlation area radar field. Then, the radar field of the strong correlation area such as the strong convective target obtained after the previous observation is gridded by using a pre-built spatiotemporal adaptive reflectivity fusion model to generate a gridded strong convective grid field of the strong correlation area. Specifically, the radar data is first processed using the spatiotemporal adaptive reflectivity fusion model. This model combines the changing characteristics of the spatiotemporal data and the radar echo intensity, and uses interpolation algorithms (such as bilinear interpolation, cubic interpolation, etc.) to convert the radar field into a regular grid structure, ensuring that each grid point represents the strong convective feature in the area, thereby generating a gridded strong convective grid field of the strong correlation area. In addition, after the gridding processing, the gridded strong convective grid field of the strong correlation area is analyzed using data mining techniques (such as support vector machines, decision trees, etc.) to extract specific strong convective feature data sets, especially for the identification of strong convective weather such as tornadoes. These features include horizontal polarization radar reflectivity. ,in is the speed of light, is the horizontal polarization electric field intensity, horizontal and vertical dual polarization differential reflectivity ,in is the horizontal polarization radar reflectivity, The radar echo correlation coefficient, which describes the correlation between horizontal and vertical polarization echoes, the horizontal and vertical dual polarization rotation angles, which describe the relative rotation angle between horizontal and vertical polarization echo signals, and the horizontal and vertical dual polarization differential phase standard deviation, which describes the fluctuation degree of the phase difference between horizontal and vertical polarization signals. Each feature is accurately measured by the radar in horizontal and vertical dual polarization modes, reflecting the radar's response characteristics to different types of severe convective weather activities, and ultimately generating a dataset of severe convective characteristics such as tornadoes.

[0072] Step S4: Obtain a preset tornado feature library, and perform tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado feature library and the tornado and other strong convective feature data set to generate a tornado-prone grid point area.

[0073] In an embodiment of the present invention, a fully verified tornado feature library is obtained, which is constructed based on historical radar data, meteorological observation data and the statistical laws of tornado occurrence. The feature library contains a variety of typical characteristics of tornadoes and severe convective weather, such as the scale, rotation speed, wind speed gradient of the tornado, and radar reflectivity, velocity spectrum, polarization difference and other information of different tornado types. A corresponding tornado recognition model is constructed to fully utilize the tornado feature library and the previously extracted tornado and other severe convective feature data sets. In this step, a supervised learning algorithm such as deep learning or support vector machine (SVM) and random forest is used to train and optimize the tornado and other severe convective recognition model through the relevant feature information extracted from the feature library and the real-time acquired radar data set. During the training process, the model will learn how to distinguish different types of tornadoes and the related severe convective weather characteristics based on the known severe convective feature data, thereby constructing a tornado and other severe convective recognition model. The dual-polarization radar technology is also used to extract and calculate the dual-polarization tornado vortex characteristic values ​​and tornado fragment characteristic values ​​of tornadoes from the strong convection identification model such as tornadoes. Through the strong convection identification model such as tornadoes, the vertical and horizontal polarization signals of the meteorological echo can be measured simultaneously, thereby obtaining detailed information about the tornado. These characteristic values ​​include but are not limited to the echo intensity, ratio, polarization difference and debris echo characteristics of the tornado. After the dual-polarization data is processed by methods such as phase difference and reflectivity difference, a series of numerical values ​​are obtained. These values ​​will be sent as input data to the tornado identification model for further analysis. By calculating these characteristic values, the rotation intensity and destructiveness of the tornado can be quantified. The dual-polarization tornado vortex characteristic values ​​include characteristic values ​​such as horizontal reflectivity measurement value, vertical reflectivity measurement value, reflectivity difference and differential phase, while the tornado fragment characteristic value is a function value composed of multiple radar parameters such as reflectivity, difference ratio and phase difference. For example, At the same time, by substituting the calculated dual-polarization tornado vortex characteristic values ​​and tornado fragment characteristic values ​​into the tornado characteristic membership function corresponding to the model, a membership metric calculation is performed. The purpose of the membership metric calculation is to compare the input characteristic values ​​with the characteristics of the tornado through a mathematical model to obtain the membership of the target area. The membership degree measures the similarity between a certain meteorological target (for example, the radar echo of a certain area) and the tornado characteristics, that is, to quantitatively calculate the similarity between the dual-polarization tornado vortex characteristic values ​​and the tornado fragment characteristic values ​​analyzed previously in the tornado characteristic library. For example, when the dual-polarization characteristic values ​​of a certain area have a high similarity with the characteristics of the tornado, the membership of the area will reach more than 90%, indicating that the possibility of a tornado occurring in the area is greater; on the contrary, if the characteristic values ​​of the area are high ...%. If the characteristics are quite different from the typical characteristics of a tornado, the membership will be lower. Then, the characteristic membership of the severe convective target tornado obtained by the above calculation is used for further radar field analysis to identify those areas with higher membership according to the membership values ​​of different areas, and to determine the grid point areas where tornadoes are most likely to occur. This analysis process combines geographic information, weather models and radar data to accurately locate areas with a higher probability of tornado occurrence (that is, tornado characteristic membership is greater than or equal to 50%). This process not only includes the analysis of the intensity distribution and velocity spectrum of the radar echo, but also involves the combination of severe convective weather characteristics (such as updrafts, low pressure, etc.) and terrain conditions to ensure that high-risk areas can be comprehensively predicted, and finally the grid point areas where tornadoes are prone to occur are identified and generated.

[0074] Further, as an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps:

[0075] Step S11: Using a micromanometer, real-time monitoring of regional micropressure in the suspicious target area is performed according to monitoring conditions corresponding to a long period of 100-250 minutes and an amplitude fluctuation greater than or equal to 60 Pa, or a short period of 20-99 minutes and an amplitude fluctuation greater than or equal to 20 Pa, to obtain micropressure change data of the suspicious area;

[0076] In an embodiment of the present invention, real-time micro-pressure monitoring of a suspicious target area is performed by using a micromanometer. The micromanometer should be set at different positions in the target area and configured to monitor long-period amplitude fluctuations with a period in the range of 100 to 250 minutes, and the amplitude of the fluctuation is not less than 60Pa; at the same time, the micromanometer must also be able to monitor short-period amplitude fluctuations in the range of 20 to 99 minutes, and the amplitude is required to be greater than or equal to 20Pa. During specific operation, the micromanometer records the micro-pressure data in the target area through periodic sampling. Within a preset time period, the micromanometer detects the pressure changes in real time according to its sensitivity and resolution, and transmits it to the data processing unit. The data processing unit determines whether there is a fluctuation pattern greater than a threshold based on the collected pressure data. If the conditions are met, it is marked as micro-pressure fluctuation data of the suspicious target area to help determine whether there are abnormal climate characteristics related to strong convection, and finally obtain the micro-pressure change data of the suspicious area.

[0077] Step S12: Using a new generation S-band weather radar, the suspicious target area is monitored for regional meteorological echoes in real time to obtain meteorological echo data of the suspicious area;

[0078] In an embodiment of the present invention, a new generation of S-band weather radar is used to perform real-time monitoring of meteorological echoes in a suspected target area. The S-band weather radar has high detection accuracy and penetration capability, and can effectively monitor meteorological echoes in the target area, especially echo signals of severe convective weather. The radar transmits electromagnetic waves and receives echo signals reflected from meteorological targets to obtain real-time meteorological change data in the target area. Especially when severe convective weather occurs, radar echoes usually appear as strong reflected signals. To obtain accurate data, the radar equipment should collect and process the reflected echo signals in real time, and determine whether there are severe convective weather characteristics in the area through different parameters (such as echo intensity, velocity gradient, etc.), ultimately obtaining meteorological echo data for the suspected area.

[0079] Step S13: performing spatiotemporal joint filtering on the micro-pressure change data of the suspicious area and the meteorological echo data of the suspicious area to obtain the micro-pressure filtered data and meteorological filtered data of the suspicious area corresponding to each spatiotemporal position point;

[0080] In an embodiment of the present invention, the micro-pressure change data and meteorological echo data of the suspicious area are subjected to spatiotemporal joint filtering processing. The purpose of this processing step is to effectively integrate the two types of data in the time and space dimensions through multi-dimensional spatiotemporal analysis, thereby removing noise and extracting relevant feature information. In specific implementation, the micro-pressure change data and meteorological echo data are first synchronized with the timestamps to ensure that the meteorological echo data and micro-pressure data at the same time can correspond. In the spatial dimension, the collected data are divided into regions and the data are assigned to corresponding spatiotemporal position points. In the process of spatiotemporal joint filtering, advanced algorithms such as Kalman filtering and particle filtering can be used. Combined with the characteristics of micro-pressure change data and radar echo data, the two are jointly filtered to optimize data quality, reduce data redundancy and errors, and obtain more accurate micro-pressure filtered data and meteorological filtered data, and finally obtain the corresponding suspicious area micro-pressure filtered data and suspicious area meteorological filtered data at each spatiotemporal position point.

[0081] Step S14: performing a matching quantitative calculation on the suspicious target area using a regional strong convection feature matching calculation formula based on the corresponding suspicious area micro-pressure filter data and the suspicious area meteorological filter data at each time and space position point to obtain a strong convection feature matching degree of the suspicious area;

[0082] In an embodiment of the present invention, a suitable feature matching calculation formula is formed by combining the micro-pressure filter data and meteorological filter data after spatiotemporal joint filtering processing, the spatial domain of the suspicious target area, the time variable parameter, the strong convection feature influence weight coefficient, the micro-pressure field gradient, the meteorological field gradient, the strong convection feature field gradient adjustment coefficient, the central reference position of the suspicious target area, the spatial position attenuation coefficient and related parameters to perform matching quantitative calculation on the suspicious target area, so as to fully consider the relationship between the micro-pressure change and the meteorological echo characteristics, such as the correlation between the fluctuation characteristics of the micro-pressure and the echo intensity, and the indicative relationship between the reflection intensity, gradient and other factors in the echo and the strong convective activity, and quantitatively calculate the strong convection feature matching degree at each spatiotemporal position point. This matching degree reflects the probability of whether there is strong convective weather in the suspicious area, and finally the strong convective feature matching degree of the suspicious area is obtained.

[0083] Step S15: Based on the strong convection feature matching degree of the suspicious area, a strong convection occurrence matching prediction is performed on the corresponding strong convection suspicious target in the suspicious target area. If the strong convection feature matching degree of the suspicious area is greater than or equal to 70%, the corresponding strong convection suspicious target in the suspicious target area is displayed, otherwise it is not displayed, so as to obtain the strong convection occurrence prediction target area.

[0084] In an embodiment of the present invention, the occurrence of the corresponding strong convective suspicious targets in the suspicious target area is predicted by combining the strong convective feature matching degree obtained by previous quantification. When the strong convective feature matching degree in the suspicious area is greater than or equal to 70%, the strong convective suspicious targets in the area are automatically displayed, prompting relevant meteorological monitoring personnel or early warning systems to pay attention to the corresponding strong convective weather risks. In specific operations, automatic judgment is made by setting a threshold. Once the matching degree exceeds 70%, the alarm mechanism will be triggered to send a strong convective prediction and warning signal to the monitoring platform or relevant personnel. If the matching degree is lower than the threshold, the strong convective target will not be displayed, indicating that strong convective weather will not occur in the area in the short term. This prediction result will be helpful for the next step of meteorological warning and decision support work, ensuring that corresponding measures are taken in time to deal with potential strong convective weather, and finally obtaining the target area for the occurrence prediction of strong convection.

[0085] Furthermore, step S12 includes the following steps:

[0086] Step S121: transmitting corresponding S-band electromagnetic waves by a new generation S-band weather radar to perform regional meteorological echo scanning on meteorological components behind the atmosphere in the suspicious target area to generate echo signals of meteorological components in the suspicious area;

[0087] In an embodiment of the present invention, a new generation of S-band weather radar is used to transmit S-band electromagnetic waves in a corresponding frequency range to scan the meteorological components behind the atmosphere in the suspicious target area. In this process, the S-band radar generates electromagnetic waves through a high-power electromagnetic wave pulse transmitter. After propagation and reflection, the radar receiver will capture the returned echo signal. This signal is mainly reflected from the clouds, water droplets and other meteorological components in the suspicious target area. The radar system ensures accurate detection of meteorological components behind the atmosphere by adjusting the transmission frequency and wavelength, reducing the impact of errors and signal attenuation. During each scan, the radar will rotate and scan at a specific angle around the predetermined area to ensure all-round echo detection of the meteorological conditions in the suspicious area, generate corresponding meteorological component echo signals, and ultimately obtain meteorological component echo signals in the suspicious area.

[0088] Step S122: analyzing the signal strength and reflectivity of the meteorological component echo signal in the suspicious area to obtain meteorological echo signal strength data and meteorological echo signal reflectivity data in the suspicious area;

[0089] In an embodiment of the present invention, after obtaining the echo signal of the meteorological component in the suspicious area, the intensity and reflectivity of the echo signal are analyzed to obtain specific signal strength data and reflectivity data, wherein the signal strength is determined by measuring the power attenuation degree of the echo signal to judge the meteorological reflection characteristics of the target area, and the reflectivity is obtained by calculating the ratio of the intensity of the echo signal reflected back to the radar receiver to the transmission intensity. The reflectivity of the echo signal is closely related to the cloud density and meteorological conditions (such as precipitation, humidity, etc.) in the suspicious area. Therefore, through reflectivity analysis, the thickness, density and other meteorological parameters of the cloud layer can be preliminarily determined, and the effective information reflecting the meteorological components in the signal can be accurately extracted, and finally the meteorological echo signal strength data and the meteorological echo signal reflectivity data of the suspicious area are obtained.

[0090] Step S123: performing a signal echo spatiotemporal distribution analysis of the meteorological components behind the atmosphere in the suspicious target area based on the suspicious area meteorological echo signal strength data and the suspicious area meteorological echo signal reflectivity data to generate a suspicious area meteorological signal echo spatiotemporal distribution map;

[0091] In an embodiment of the present invention, a spatiotemporal distribution analysis of the signal echo is performed based on the obtained meteorological echo signal strength data and reflectivity data of the suspicious area. The core of this step is to process the spatiotemporal changes of the echo signal data to analyze the spatial and temporal distribution patterns of the meteorological components. The radar system needs to use high-speed data acquisition and processing equipment to record the signal strength and reflectivity data in chronological order, and use spatiotemporal analysis algorithms, such as time series models and spatial interpolation methods, to model the spatiotemporal distribution of the meteorological echo data in the suspicious area. This process can analyze the temporal changes of the echo signal (for example, the change of echo intensity over time) and combine it with spatial geographic information system (GIS) technology to generate a corresponding spatiotemporal distribution map. The map reflects the dynamic changes of the meteorological components and ultimately generates a spatiotemporal distribution map of the meteorological signal echo in the suspicious area.

[0092] Step S124: performing a meteorological spatiotemporal distribution field simulation analysis on the suspicious target area according to the spatiotemporal distribution map of meteorological signal echoes in the suspicious area to generate a spatiotemporal distribution field of meteorological echoes in the suspicious area;

[0093] In an embodiment of the present invention, a simulation analysis of the meteorological spatiotemporal distribution field of the corresponding suspicious target area is performed based on the spatiotemporal distribution map of the meteorological signal echo in the suspicious area. In this step, a numerical simulation system combining a physical model with meteorological data is established to simulate the meteorological spatiotemporal distribution field of the suspicious area. The models used include atmospheric dynamics equations and turbulent diffusion models to simulate the movement and distribution of meteorological components. In a specific implementation, the finite element analysis method (FEM) or the finite difference method (FDM) can be used to interpolate and fit the data in the spatiotemporal distribution map of the meteorological echo to construct a three-dimensional meteorological spatiotemporal distribution field. The input data of the model comes from the previously generated spatiotemporal distribution map, and factors such as changes in atmospheric temperature and humidity, wind speed and direction are taken into account, thereby forming a spatiotemporal distribution field of meteorological echoes in the suspicious area, and finally generating a spatiotemporal distribution field of meteorological echoes in the suspicious area.

[0094] Step S125: Real-time monitoring of the spatiotemporal distribution of meteorological echoes in the suspicious area is performed to obtain meteorological echo data in the suspicious area, wherein the meteorological echo data in the suspicious area includes corresponding meteorological conditions and cloud density changes in the suspicious target area.

[0095] In an embodiment of the present invention, by real-time monitoring of the spatiotemporal distribution field of meteorological echoes in suspicious areas, a radar-based real-time meteorological monitoring system is established to continuously collect and analyze meteorological echo signals in suspicious areas. In this process, the equipment used includes a real-time data processing server and a cloud computing platform for storing and analyzing large-scale meteorological data. The meteorological echo data at this stage not only includes the intensity and reflectivity of the echo signal, but also needs to be comprehensively analyzed in combination with other meteorological parameters such as cloud density, temperature, and humidity. Through the real-time monitoring system, the collected data is quickly transmitted to the processing platform, processed within a few seconds and a visual chart is generated to reflect information such as cloud changes and wind field distribution. The real-time monitoring system needs to cooperate with a high-precision meteorological early warning algorithm to respond quickly to abnormal changes, prompt the formation and evolution of corresponding extreme weather such as tornadoes, and finally obtain meteorological echo data of suspicious areas, including the corresponding meteorological conditions and cloud density changes in the suspicious target area.

[0096] Furthermore, the regional severe convection feature matching calculation formula in step S14 is specifically:

[0097] ;

[0098] Where, For suspicious target areas at time and location points The matching degree of strong convection characteristics in the corresponding suspicious area is: is the spatial domain of the suspicious target area, is the start time of the time range, is the end time of the time range, For suspicious target areas at time and location points The corresponding micro-pressure parameters of the suspicious area are: For suspicious target areas at time and location points The corresponding meteorological parameters of the suspicious area are as follows: is the characteristic parameter field of severe convection in the suspicious target area, is the weight coefficient of the strong convection characteristics, For suspicious target areas at time and location points The corresponding micro-pressure field gradient is: For suspicious target areas at time and location points The corresponding meteorological field gradient is is the gradient adjustment coefficient of the strong convection characteristic field, is an exponential function, is the reference position of the center of the suspicious target area, is the spatial position attenuation coefficient, is the correction coefficient for the matching degree of severe convective characteristics in the suspicious area.

[0099] The present invention obtains a regional severe convection feature matching calculation formula by using a specific mathematical model and after verification, which is used to perform matching quantitative calculations on suspicious target areas. The regional severe convection feature matching calculation formula can effectively capture the potential characteristics of severe convective weather by combining micro-pressure change data and meteorological data and performing comprehensive analysis based on gradients. The gradient changes of micro-pressure fields and meteorological fields usually change significantly during the formation and evolution of severe convective weather. Therefore, combining micro-pressure and meteorological echo data for joint filtering and feature matching can enhance the accurate identification capability of severe convective weather events. The relevant parameters in this formula are: and It can be adjusted according to actual conditions to quantitatively reflect the effects of different influencing factors on the characteristics of severe convection, for example, represents the influence weight of strong convective characteristics, Adjusting the gradient effect of the meteorological field and the micro-pressure field can further optimize the calculation results and improve the accuracy of the overall matching. In the calculation formula, the introduction of the spatial position attenuation coefficient can effectively reflect the phenomenon that the influence of positions farther away from the center of the target area gradually weakens. The spatial attenuation model is usually used to describe the law that certain physical quantities in the geographic space weaken as the distance increases. Therefore, for the prediction of severe convection, it can more accurately identify severe convective events located near the center of the target area, thereby avoiding the influence of areas far from the center on the final prediction results. In addition, by introducing the correction coefficient, the formula can adaptively adjust the matching calculation results during actual operation, and increase the prediction ability of sudden or nonlinear severe convective events. For example, if there are atypical meteorological conditions or sudden changes, the correction coefficient can help improve the prediction calculation process so that the prediction results are more consistent with the actual severe convective events. In summary, the formula fully takes into account the suspicious target area at time and location points The matching degree of the strong convection characteristics of the corresponding suspicious area , the spatial domain of the suspicious target area , the time range starts at , the time range ends , the suspicious target area at time and location points The corresponding micro-pressure parameters of the suspicious area , the suspicious target area at time and location points The corresponding meteorological parameters of the suspicious area , characteristic parameter field of severe convection in suspicious target area , the weight coefficient of strong convection characteristics , the suspicious target area at time and location points The corresponding micro-pressure field gradient , the suspicious target area at time and location points The corresponding meteorological field gradient , gradient adjustment coefficient of strong convection characteristic field , exponential function , the reference position of the center of the suspicious target area , spatial position attenuation coefficient , correction coefficient of the matching degree of strong convective characteristics in the suspicious area , according to the suspicious target area at time and location points The matching degree of strong convection characteristics in the corresponding suspicious area The mutual correlation between the above parameters constitutes a functional relationship:

[0100] ;

[0101] This formula can realize the quantitative calculation process of matching the suspicious target area, and at the same time, the correction coefficient of the matching degree of the strong convective characteristics of the suspicious area is used. The introduction of can be adjusted according to the errors that occur during the calculation process, thereby improving the accuracy and applicability of the calculation formula for regional severe convective feature matching.

[0102] Furthermore, step S2 includes the following steps:

[0103] Step S21: obtaining the solar radio radiation intensity at the corresponding signal wavelength of the corresponding radar detection equipment in the tornado detection network under the current solar radiation;

[0104] In an embodiment of the present invention, by accurately measuring the impact of solar radiation on radar signals, real-time data of solar radiation intensity is first obtained. During specific implementation, it is necessary to rely on solar radio radiation models and related instruments to measure solar radiation intensity. This can be achieved by setting up solar radiation monitoring equipment, such as solar radiation detection instruments, in the tornado detection network to obtain the solar radiation intensity at the current time and geographical location in real time. The solar radio radiation intensity model is used to calculate the solar radio radiation intensity corresponding to the radar signal wavelength based on the spectrum of solar radiation and the signal wavelength used by the radar equipment. This process requires accurate frequency characteristics of the radiation intensity and is adjusted in combination with factors such as the solar activity cycle to obtain the radio radiation intensity suitable for the current conditions, and finally obtain the solar radio radiation intensity of the radar detection equipment at the corresponding signal wavelength under the current solar radiation.

[0105] Step S22: Calculate the received signal power of the corresponding radar detection equipment in the tornado detection network based on the solar radio radiation intensity at the corresponding signal wavelength under the current solar radiation to obtain the theoretical value of the radar detection received signal power. The calculation formula of the theoretical value of the radar detection received signal power is specifically: ;

[0106] in, is the theoretical value of radar detection received signal power, is the radar receiving frequency bandwidth, is the radar antenna gain, is the radar signal wavelength, The wavelength of the radar signal The intensity of solar radio radiation at The radar receiving signal frequency;

[0107] In an embodiment of the present invention, a calculation formula is formed by combining the radar receiving frequency bandwidth, radar antenna gain, radar signal wavelength, solar radio radiation intensity at the corresponding signal wavelength under current solar radiation, radar receiving signal frequency and related parameters to calculate the receiving signal power of the corresponding radar detection equipment in the tornado detection network, so as to quantify the theoretical power value of the radar receiving signal and finally obtain the theoretical value of the radar detection receiving signal power.

[0108] Step S23: obtaining a measured value of the radar detection received signal power, and performing a power difference calculation based on the measured value of the radar detection received signal power and a theoretical value of the radar detection received signal power to obtain a radar detection received signal power difference;

[0109] In an embodiment of the present invention, actual signal reception power is measured by each radar device within the tornado detection network. This process is monitored in real time by a power meter installed on the radar detection device to obtain the actual received signal power value. These measured values ​​reflect the signal power received by the radar device in actual operation. After obtaining the measured values, they are compared with the theoretical signal power previously calculated by solar radio nuclei to calculate the power difference. ,in is the power difference, is the measured value of the radar detection received signal power, is the theoretical value of radar detection received signal power, and finally the radar detection received signal power difference is obtained.

[0110] Step S24: performing solar radio calibration processing on the corresponding radar detection equipment in the tornado detection network based on the difference in radar detection received signal power to generate a tornado detection calibration optimization network;

[0111] In an embodiment of the present invention, solar radio calibration is performed on the corresponding radar devices in the tornado detection network by utilizing the previously calculated radar detection received signal power differences. In specific implementation, the deviation value of each device is first calculated based on the power difference of each radar device. This deviation value can reflect the impact of solar radio radiation on different radar devices. Based on these deviation values, the radar equipment is optimized by adjusting parameters such as the radar's receiving sensitivity and gain. This process can be performed through an automatic calibration system, which includes automated calibration software and hardware interfaces, and can fine-tune the radar equipment in real time so that the radar's received signal power is more consistent with the theoretical predicted value. By calibrating all radar devices, an optimized detection network that can more accurately respond to solar radiation interference is formed, and finally a tornado detection calibration optimization network is generated.

[0112] Step S25: triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspicious target within the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspicious target within the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target.

[0113] In an embodiment of the present invention, a tornado detection strategy is triggered based on real-time weather data and radar observation results in the severe convection prediction area, and through the meteorological warning system and the severe convection prediction model, potential severe convection suspected target areas are first identified, and this information is used as input to trigger the corresponding detection strategy. Next, a collaborative detection instruction is started according to the optimized tornado detection calibration optimization network, and multiple radar devices are directed to perform collaborative observation based on the position, speed and direction of the target area. This process transmits collaborative instructions through a real-time data link, allowing multiple radar devices to work synchronously, thereby forming a three-dimensional radar field covering the target area. The three-dimensional radar field can provide multi-dimensional monitoring of severe convection targets, including information such as the target's height, speed and spatial distribution, and finally observes and generates a three-dimensional radar field of severe convection targets.

[0114] Furthermore, step S25 includes the following steps:

[0115] Step S251: triggering the built-in policy control server to respond and send a tornado detection policy coordination instruction based on the existence of a suspected strong convection target corresponding to the strong convection prediction target area;

[0116] In an embodiment of the present invention, suspected targets where tornadoes may occur are identified by predicting target areas for occurrence of severe convection based on previous predictions. These targets are obtained through algorithm analysis, and their probability of occurrence is determined based on meteorological monitoring data such as wind speed, air pressure changes, temperature and humidity, and other parameters. The prediction results are analyzed by a policy control server, and the built-in policy system is used to determine which areas have a higher risk of tornado occurrence. After identifying a severe convection area with potential threats, the control server will immediately generate and send a collaborative instruction to start a tornado detection task. The instruction clearly specifies the target area to be detected, the predetermined detection equipment, and the predetermined task plan, ensuring that each collaborative detection system can accurately cooperate according to the instruction, start the corresponding detection work, and finally respond to send a tornado detection strategy collaborative instruction.

[0117] Step S252: applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network and generating a tornado detection coordination scanning command in response;

[0118] In an embodiment of the present invention, the tornado detection strategy coordination instruction generated by the previous response is transmitted to the tornado detection calibration optimization network through the communication network. The network includes a search radar, an optical camera and an ultra-fine radar. According to the content of the coordination instruction, the equipment in the network will optimize the calibration to ensure that each detection device can accurately carry out the task according to the set scanning mode at the same time. The calibration optimization process ensures the accuracy and consistency of the detection results by correcting factors such as the time synchronization, spatial position error and signal transmission delay of the equipment. After optimization, a tornado detection coordinated scanning command is generated and the command is fed back to each detection device. These devices will adjust their scanning strategies according to the scanning parameters set by the instruction to ensure all-round coverage and efficient detection of the strong convection target area, and finally respond to generate a tornado detection coordinated scanning command.

[0119] Step S253: applying the tornado detection collaborative scanning command to the corresponding search radar in the tornado detection calibration optimization network to perform target area detection scanning on the corresponding strong convection suspected targets within the strong convection occurrence prediction target area, so as to generate the target scanning signal echo intensity corresponding to each strong convection occurrence target area;

[0120] In an embodiment of the present invention, by using an optimized calibration network instruction, it will be sent to a specific radar device, especially a search radar that scans a strong convection area. The search radar adjusts its working mode according to the collaborative scanning command and begins to accurately detect the predicted target area. The radar system continuously or intermittently scans the target area by emitting electromagnetic wave signals and receives echo signals in real time. The intensity of the echo signal is closely related to the size, density and other characteristics (such as precipitation intensity, airflow disturbance, etc.) of the reflected target object. By analyzing the echo signal, the radar can extract the echo intensity of each strong convection target area, thereby providing basic data for subsequent target identification, positioning and tracking, ensuring the accurate collection of echo intensity data for each strong convection target area, and finally scanning to generate the target scanning signal echo intensity corresponding to each strong convection target area.

[0121] Step S254: applying the tornado detection collaborative scanning command to the corresponding optical camera in the tornado detection calibration optimization network to perform target synchronous acquisition scanning on the corresponding strong convection suspected targets in the strong convection prediction target area to generate the target area size and target area position corresponding to each strong convection target area; calculating the scanning azimuth angle range based on the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convection target area to generate the strong convection target area scanning azimuth angle distribution range;

[0122] In the embodiment of the present application, the tornado detection cooperative scanning command is transmitted to the optical camera instrument in the tornado detection calibration optimization network to perform synchronous acquisition scanning of the target area. The optical camera instrument captures visual images of the strong convective target area in real time through a high-resolution camera and a video acquisition device. These images can be used to verify radar echo data and provide more accurate target area size and spatial position. Image processing technology is used to extract the boundaries of the target area from the acquired images to further determine the length, width, area, and center position of the target area, thereby generating the target area size and target area position corresponding to each strong convective target area. At the same time, by combining these information with the target scanning signal echo intensity, the spatial distribution and dynamic change of the target area are further analyzed. Subsequently, using the echo intensity, area size, and position data of the target area, the scanning range of the target area at different azimuth angles is calculated, and by accurately measuring the geometry and position of the target area, the scanning azimuth angle distribution range is generated to ensure that all key parts of the target area are fully covered, and finally the strong convective target area scanning azimuth angle distribution range is generated.

[0123] Step S255: The tornado detection cooperative scanning command is applied to the corresponding hyperfine radar in the tornado detection calibration optimization network, and based on the strong convective target area scanning azimuth angle distribution range, the strong convective suspicious target in the strong convective prediction target area is fixed-point cooperative sector scanning tracked and three-dimensionally analyzed to generate a strong convective target three-dimensional radar field.

[0124] In the embodiment of the present application, the hyperfine radar device starts fixed-point sector scanning tracking of the strong convective target area according to the tornado detection cooperative scanning command. At this time, the hyperfine radar system controls the angle by the strong convective target area scanning azimuth angle distribution range measured previously, focuses on the key position of the strong convective target area, and performs high-precision scanning. Fixed-point sector scanning enables the radar to continuously track the changes of the target and update the radar image data of the target area in real time. At the same time, based on the scanning azimuth angle distribution range of the target area, the radar adjusts the direction and range of the scanning area, thereby improving the capture ability of dynamic targets. On this basis, the radar system further performs three-dimensional analysis of the target area to construct a three-dimensional radar field of the strong convective target area through scanning data at different height levels. This radar field provides detailed spatial distribution and intensity information of the target area, which can reflect the three-dimensional structure, motion trajectory, and airflow pattern of the target area, and finally generates a strong convective target three-dimensional radar field.

[0125] Further, the scanning azimuth angle range calculation in step S254 according to the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convective target area includes the following steps:

[0126] Performing spatial coordinate system conversion according to the target area position corresponding to each severe convection target area to generate a severe convection target area position spatial coordinate system;

[0127] In an embodiment of the present invention, the target area position of each target area where severe convection occurs is obtained, and the position is usually represented by the latitude and longitude coordinates obtained by the radar system through measurement. In actual operation, the accurate geographic coordinates of the target area can be determined through the spatiotemporal information of the radar echo signal, combined with the latitude and longitude coordinates of the target area and the position and scanning angle information of the radar. These geographic coordinates need to be converted into a spatial coordinate system suitable for subsequent analysis, for example, a local plane coordinate system or a polar coordinate system related to the radar platform. In the coordinate conversion process, a spatial conversion matrix is ​​used (such as converting the geographic coordinate system into the radar coordinate system), and the geographic location of the target area is mapped to the new coordinate system through rotation and offset calculations in three-dimensional space. At this time, the spatial position of the severe convection target area is described in a standard coordinate system, and finally a spatial coordinate system of the position of the severe convection target area is generated.

[0128] Preferably, the target area size corresponding to each strong convection target area is measured and calculated based on the spatial coordinate system of the strong convection target area position, and the regional scale characteristic index corresponding to each strong convection target area is obtained, wherein the regional scale characteristic index includes the regional major axis, the regional minor axis and the total area size of the region;

[0129] In an embodiment of the present invention, regional scale measurement is performed by combining a previously determined target region spatial coordinate system and calculating the boundary of the severe convective target region. The boundary of the region is usually determined by multiple radar echo reflection intensity data points. These points form a polygon or closed curve, representing the outer contour of the severe convective target region. Using a geometric algorithm (such as a minimum circumscribed rectangle or a minimum enclosing ellipse), the major axis and minor axis of the region can be fitted through a set of discrete boundary points. In this process, by calculating the distribution of all data points in the region, the least squares method is used to fit the optimal ellipse or rectangle to determine the major axis, minor axis and total area size. The major axis of the region is the length of the straight line representing the maximum expansion direction of the region, the minor axis is the shortest length in the direction perpendicular to the major axis, and the total area of ​​the region is the area size of the fitted shape. These scale characteristic indicators provide quantitative physical characteristics for subsequent regional analysis, scanning angle calculation and dynamic tracking of severe convective targets, and ultimately obtain regional scale characteristic indicators corresponding to each severe convective target region.

[0130] Preferably, the scanning azimuth angle range is calculated for the target scanning signal echo intensity and regional scale characteristic index corresponding to each strong convection target area based on the strong convection target area position spatial coordinate system to generate the strong convection target area scanning azimuth angle distribution range.

[0131] In an embodiment of the present invention, by utilizing the previously obtained regional scale characteristic index and spatial coordinate system, the radar scanning signal echo intensity data is combined with the positional relationship of the target area to calculate the scanning azimuth angle range. The radar scanning signal intensity and the boundary data of the target area can be used to infer the angular distribution range of the strong convection target area in the radar scan. By setting the starting angle and ending angle of the scan, and based on the position, direction and scale characteristics (such as the long axis direction) of the target area, the echo intensity distribution of the target area at each azimuth angle can be inferred. Specifically, the long axis direction of the area is used to determine the main direction of the scan, and combined with the vertical direction of the radar scan, the signal reflection range of the target area at different scanning angles is calculated. This calculation process is usually performed using a polar coordinate system, where the scanning angle range is determined by the radar scanning angle and the spatial position relationship of the target area. The obtained scanning azimuth angle range is the spatial projection of the strong convection target area in the radar scan, and finally the scanning azimuth angle distribution range of the strong convection target area is generated.

[0132] Furthermore, step S3 includes the following steps:

[0133] Step S31: obtaining a severe convective weather activity characteristic data matrix and an environmental condition distribution climate characteristic data matrix corresponding to the severe convective suspicious target in the severe convective target three-dimensional radar field;

[0134] In an embodiment of the present invention, relevant data of suspected strong convective targets are obtained through the three-dimensional radar field of strong convective targets obtained through previous observations. These radar data are converted into a strong convective weather activity characteristic data matrix, which reflects the intensity, distribution characteristics and change pattern of the radar echo signal. At the same time, the environmental condition distribution climate characteristic data matrix is ​​also collected, which contains multi-dimensional environmental data such as temperature, humidity, air pressure, wind speed, and wind direction. These environmental condition data can be obtained through automatic weather stations, meteorological satellites and other meteorological detection equipment, and correspond to the radar data. In the radar signal analysis, the echo intensity, speed, kinetic energy and other information returned by the radar are mainly analyzed. These data are extracted and stored as a characteristic data matrix through a preset algorithm for subsequent processing, and finally the strong convective weather activity characteristic data matrix and the environmental condition distribution climate characteristic data matrix are obtained.

[0135] Step S32: Based on the severe convective weather activity characteristic data matrix and the environmental condition distribution climate characteristic data matrix, a canonical correlation analysis expression is used to perform a canonical correlation calculation on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field to obtain a canonical correlation coefficient between the strong convective weather activity corresponding to the strong convective suspicious targets and the environmental conditions;

[0136] In an embodiment of the present invention, the correlation between the characteristics of severe convective weather activities and the characteristics of environmental conditions is calculated by the canonical correlation analysis (CCA) method. Canonical correlation analysis is a multivariate statistical method that aims to reveal the association pattern between the weather activity characteristic data matrix and the environmental condition distribution climate characteristic data matrix by finding the maximum correlation between two groups of variables. In the specific operation, the severe convective weather activity characteristic data matrix and the environmental condition data matrix are used as input, and after standardization, the canonical correlation analysis expression is used for calculation. Through calculation, the typical variables of each of the two matrices are obtained, and then the canonical correlation coefficient is calculated to reflect the intrinsic connection between severe convective weather and environmental conditions. Specifically, using the calculated canonical correlation coefficient, it is possible to evaluate which environmental conditions have a greater impact on severe convective weather activities, thereby helping to identify and classify suspected severe convective targets, and finally obtaining the canonical correlation coefficient between the severe convective weather activities and environmental conditions corresponding to the suspected severe convective targets.

[0137] Among them, the canonical correlation analysis expression is specifically:

[0138] ;

[0139] Where, is the canonical correlation coefficient, is the coefficient vector corresponding to the typical variables in the characteristic data matrix of severe convective weather activities, is the coefficient vector corresponding to the typical variables in the climate characteristic data matrix of environmental condition distribution, is the vector transpose symbol, is the characteristic data matrix of severe convective weather activities, is the climate characteristic data matrix of environmental conditions distribution, is the expected value, is the covariance matrix between severe convective weather activities and environmental conditions, is the autocovariance matrix of severe convective weather activity data, is the autocovariance matrix of environmental condition data;

[0140] Step S33: performing target equal-correlation classification on the corresponding strong convective suspicious targets in the three-dimensional radar field based on the typical correlation coefficient between the strong convective weather activities corresponding to the strong convective suspicious targets and the environmental conditions, so as to obtain a set of strong convective target equal-correlation;

[0141] In an embodiment of the present invention, correlation classification of strong convective targets corresponding to suspicious strong convective targets in a three-dimensional radar field of strong convective targets is performed based on a typical correlation coefficient obtained by previous quantitative calculation. First, cluster analysis is performed on the suspicious strong convective targets using the previously calculated typical correlation coefficient, so as to divide the targets into different correlation categories by setting a certain correlation threshold. Targets with typical correlation coefficients higher than the threshold are classified into the same strong correlation category, indicating that the correlation between the strong convective weather activities and environmental conditions corresponding to these targets is strong; targets with lower correlation are classified into weak correlation categories. Specifically, the calculated typical correlation coefficient is processed using a clustering algorithm in machine learning (such as K-means, DBSCAN, etc.), and the suspicious strong convective targets in the radar field are automatically classified into equal-strong correlation categories. Each category of targets represents an area in the radar field with similar weather activity and environmental condition characteristics, and finally a set of equal-strong correlations of strong convective targets is obtained.

[0142] Step S34: performing target strong correlation region fusion division on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field based on the strong convective target equal strong correlation set to generate a strong convective target equal strong correlation region radar field;

[0143] In an embodiment of the present invention, strong convective target regions are divided by combining and aggregating strong convective target equal-correlation sets corresponding to strong convective suspicious targets in the three-dimensional radar field of strong convective targets, so as to realize spatial distribution analysis of targets in each equal-correlation category, and aggregate targets with the same strong correlation into one region. In order to ensure the accuracy of the division, spatial interpolation methods (such as Kriging interpolation, spline interpolation, etc.) can be used to smooth the spatial distribution of strong convective targets, and adjacent equal-correlation targets can be merged into larger regions through spatial clustering technology. In this process, factors such as the target's geographical location, spatial characteristics of radar echoes, and environmental conditions are all used as inputs to ensure that the generated strong convective target equal-correlation region radar field can accurately reflect the distribution of meteorological activities in different regions. The generated radar field represents the area where strong convective weather activities and environmental conditions are similar in space, and finally a strong convective target equal-correlation region radar field is generated.

[0144] Step S35: Gridding the radar field of the strong convective target and other strong correlation areas using the spatiotemporal adaptive reflectivity fusion model to generate a gridded strong convective grid field of the strong correlation areas; performing strong convective feature analysis on the gridded strong convective grid field of the strong correlation areas to generate a tornado and other strong convective feature dataset, wherein the tornado and other strong convective feature dataset includes horizontal polarization radar reflectivity. , horizontal and vertical dual polarization differential reflectivity , radar echo correlation coefficient , horizontal and vertical dual polarization rotation angle And the horizontal and vertical dual polarization differential phase standard deviation ,in is the differential phase measured by the radar under horizontal and vertical dual polarization.

[0145] In an embodiment of the present invention, a pre-built spatiotemporal adaptive reflectivity fusion model is used to perform corresponding gridding processing on the radar field of a strong correlation area such as a strong convective target obtained after previous observation, so as to generate a gridded strong convection grid field of the strong correlation area. Specifically, the spatiotemporal adaptive reflectivity fusion model is first used to process the radar data. The model converts the radar field into a regular grid structure by combining the changing characteristics of the spatiotemporal data and the radar echo intensity using an interpolation algorithm (such as bilinear interpolation, cubic interpolation, etc.), ensuring that each grid point represents the strong convective characteristics in the area. The expression corresponding to the spatiotemporal adaptive reflectivity fusion model is specifically: ;

[0146] Where, Output value of the spatiotemporal adaptive reflectivity fusion model, indicating the center position of the grid and time The gridded strong convective reflectivity value at is the maximum number of grid points, is the item index of the horizontal grid point, is the item index of the vertical grid, Horizontal grid With vertical grid The weight coefficient matrix between For the The horizontal position parameters corresponding to the horizontal grid points, For the The vertical position parameters corresponding to the vertical grid points, At the grid point and time The original radar observed reflectivity value at At the grid point and the previous moment The original radar observed reflectivity value at At the grid point and the previous moment The gridded strong convection reflectivity values ​​at , thereby generating a gridded strong convection grid field of equal-strong correlation areas. At the same time, after the gridding process, data mining techniques (such as support vector machine, decision tree, etc.) are used to perform feature analysis on the gridded strong convection grid field of equal-strong correlation areas to extract specific strong convection feature data sets, especially for the identification of strong convective weather such as tornadoes. These features include horizontal polarization radar reflectivity, ,in is the speed of light, is the horizontal polarization electric field intensity, horizontal and vertical dual polarization differential reflectivity ,in is the horizontal polarization radar reflectivity, The vertical polarization radar reflectivity, radar echo correlation coefficient, which describes the correlation between horizontal and vertical polarization echoes, horizontal and vertical dual polarization rotation angles, which are the relative rotation angles between horizontal and vertical polarization echo signals, and horizontal and vertical dual polarization differential phase standard deviations, which are the fluctuation degree of the phase difference between horizontal and vertical polarization signals. Each feature is accurately measured by the radar in horizontal and vertical dual polarization modes, reflecting the radar response characteristics to different types of severe convective weather activities, and finally generating a dataset of severe convective features such as tornadoes, including horizontal polarization radar reflectivity. , horizontal and vertical dual polarization differential reflectivity , radar echo correlation coefficient , horizontal and vertical dual polarization rotation angle And the horizontal and vertical dual polarization differential phase standard deviation ,in is the differential phase measured by the radar under horizontal and vertical dual polarization.

[0147] Furthermore, step S4 includes the following steps:

[0148] Step S41: obtaining a preset tornado feature library;

[0149] In an embodiment of the present invention, a fully verified tornado feature library is obtained. The feature library is constructed based on historical radar data, meteorological observation data, and the statistical laws of tornado occurrence. The feature library contains typical characteristics of various tornadoes and severe convective weather, such as the scale, rotation speed, wind speed gradient of the tornado, and radar reflectivity, velocity spectrum, polarization differences, and other information of different tornado types. This feature library is optimized by analyzing radar data of a large number of historical weather events, and taking into account the climate change characteristics of different geographical regions and seasonality. The feature library is established using data mining technology and machine learning algorithms. Model training is performed through a large amount of labeled data, so that the library can accurately identify the basic characteristics of various types of tornadoes, and ultimately obtain a tornado feature library.

[0150] Step S42: Construct a tornado and other strong convection identification model based on the tornado feature library and the tornado and other strong convection feature data set, and calculate the dual-polarization tornado vortex characteristic value based on the tornado and other strong convection identification model and tornado debris characteristic values ;

[0151] In an embodiment of the present invention, a corresponding tornado recognition model is constructed to fully utilize the tornado feature library and the previously extracted tornado and other severe convection feature dataset. In this step, a supervised learning algorithm such as deep learning or support vector machine (SVM) and random forest is used to train and optimize the tornado and other severe convection recognition model through the relevant feature information extracted from the feature library and the real-time radar dataset. Specifically, the model will include multiple processing modules, such as a data preprocessing module (for removing noise and outliers), a feature extraction module (for extracting effective dual-polarization radar features, wind speed data and climate data from the original radar data), and a prediction module (for extracting effective dual-polarization radar features, wind speed data and climate data from the original radar data). During the training process, the model will learn how to distinguish different types of tornadoes and related severe convective weather characteristics based on known severe convective feature data. The model continuously adjusts parameters based on the evaluation results to improve recognition accuracy. Appropriate algorithms and tools should be selected during the model training process, such as TensorFlow, PyTorch and other frameworks to build neural networks, or use traditional machine learning algorithms for feature classification to build a severe convection recognition model for tornadoes and other severe convection. At the same time, by using radar dual-polarization technology, the dual-polarization tornado vortex characteristic values ​​and tornado fragment characteristic values ​​of tornadoes are extracted and calculated from strong convection identification models such as tornadoes. Through strong convection identification models such as tornadoes, the vertical and horizontal polarization signals of meteorological echoes can be measured simultaneously, thereby obtaining detailed information about tornadoes, such as vortex intensity, size, structure and fragment distribution characteristics. These characteristic values ​​include but are not limited to the echo intensity, ratio, polarization difference and fragment echo characteristics of tornadoes. After the dual-polarization data is processed by methods such as phase difference and reflectivity difference, a series of numerical values ​​are obtained. These values ​​will be sent as input data to the tornado identification model for further analysis. By calculating these characteristic values, the rotation intensity and destructiveness of the tornado can be quantified, and finally the dual-polarization tornado vortex characteristic values ​​are obtained. and tornado debris characteristic values , where the dual-polarization tornado vortex characteristic values ​​include horizontal reflectivity measurement value, vertical reflectivity measurement value, reflectivity difference and differential phase and other characteristic values, while the tornado debris characteristic value is a function value composed of multiple radar parameters such as reflectivity, difference ratio and phase difference, for example, .

[0152] Step S43: The dual polarized tornado vortex characteristic value and tornado debris characteristic values Substitute the tornado characteristic membership function corresponding to the severe convection identification model such as tornado to perform membership metric calculation to obtain the severe convection target tornado characteristic membership degree;

[0153] In an embodiment of the present invention, the calculated dual-polarized tornado vortex characteristic value and tornado fragment characteristic value are substituted into the tornado characteristic membership function corresponding to the model to perform a membership metric calculation. The purpose of the membership metric calculation is to compare the input characteristic value with the tornado characteristic through a mathematical model to obtain the membership degree of the target area. The membership degree measures the similarity between a certain meteorological target (for example, the radar echo of a certain area) and the tornado characteristic, that is, to quantitatively calculate the similarity between the dual-polarized tornado vortex characteristic value and the tornado fragment characteristic value analyzed previously in the tornado characteristic library. The function usually adopts the fuzzy logic method to classify each meteorological data point through the fuzzy membership function. For example, when the bipolarization characteristic value of a certain area is highly similar to the characteristics of a tornado, the membership degree of the area will reach more than 90%, indicating that the possibility of a tornado occurring in the area is greater; conversely, if the characteristic value of the area is significantly different from the typical characteristics of a tornado, the membership degree will be lower. During the calculation process, the membership function in the fuzzy inference system or neural network can be used to complete the calculation task to ensure that the results can reflect the actual weather characteristics and ultimately obtain the characteristic membership degree of the severe convective target tornado.

[0154] Step S44: performing tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado characteristic membership of the strong convective target to generate a grid point area prone to tornado occurrence.

[0155] In an embodiment of the present invention, the characteristic membership of the severe convective target tornado obtained by the aforementioned calculation is used to perform further radar field analysis to identify areas with higher membership values ​​based on the membership values ​​of different regions, thereby determining the grid points where tornadoes are most likely to occur. This analysis process combines geographic information, weather models, and radar data to accurately locate areas with a higher probability of tornado occurrence (i.e., tornado characteristic membership greater than or equal to 50%). This process not only includes analysis of the intensity distribution and velocity spectrum of radar echoes, but also involves the combination of severe convective weather characteristics (such as updrafts, low pressure, etc.) and terrain conditions to ensure that high-risk areas can be comprehensively predicted and ultimately identify grid points where tornadoes are prone to occur.

[0156] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A tornado collaborative observation method based on detection equipment, characterized in that: The detection equipment includes a micromanometer, an S-band new generation weather radar, a search radar, an optical camera, and an ultra-fine radar, wherein the search radar is connected to the optical camera and the ultra-fine radar respectively, and the optical camera and the ultra-fine radar are connected to form a tornado detection network corresponding to a triangular connection relationship. The tornado detection network is electrically connected to the micromanometer and the S-band new generation weather radar respectively. The tornado collaborative observation method based on the detection equipment includes the following steps: Step S1: Using a micromanometer and a new generation S-band weather radar, the suspected target area is monitored for regional micropressure and meteorological echoes in real time to obtain micropressure change data and meteorological echo data in the suspected area; based on the micropressure change data and meteorological echo data in the suspected area, a strong convection occurrence matching prediction is performed on the suspected target area to obtain a predicted target area for strong convection occurrence; Step S2: performing solar radio calibration on the corresponding radar detection equipment in the tornado detection network to generate a tornado detection calibration optimization network; triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspected target in the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspected target in the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target; Step S3: performing target strong correlation region fusion division on the corresponding strong convective suspicious targets in the three-dimensional radar field of the strong convective target to generate a radar field of strong convective target and strong correlation region; performing isotropic convection feature analysis on the radar field of strong convective target and strong correlation region using a spatiotemporal adaptive reflectivity fusion model to generate a dataset of strong convective features such as tornadoes; Step S4: Obtain a preset tornado feature library, and perform tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado feature library and the tornado and other strong convective feature data set to generate a tornado-prone grid point area.

2. The tornado collaborative observation method based on detection equipment according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Using a micromanometer, real-time monitoring of regional micropressure in the suspicious target area is performed according to monitoring conditions corresponding to a long period of 100-250 minutes and an amplitude fluctuation greater than or equal to 60 Pa, or a short period of 20-99 minutes and an amplitude fluctuation greater than or equal to 20 Pa, to obtain micropressure change data of the suspicious area; Step S12: Using a new generation S-band weather radar, the suspicious target area is monitored for regional meteorological echoes in real time to obtain meteorological echo data of the suspicious area; Step S13: performing spatiotemporal joint filtering on the micro-pressure change data of the suspicious area and the meteorological echo data of the suspicious area to obtain the micro-pressure filtered data and meteorological filtered data of the suspicious area corresponding to each spatiotemporal position point; Step S14: performing a matching quantitative calculation on the suspicious target area using a regional strong convection feature matching calculation formula based on the corresponding suspicious area micro-pressure filter data and the suspicious area meteorological filter data at each time and space position point to obtain a strong convection feature matching degree of the suspicious area; Step S15: Based on the strong convection feature matching degree of the suspicious area, a strong convection occurrence matching prediction is performed on the corresponding strong convection suspicious target in the suspicious target area. If the strong convection feature matching degree of the suspicious area is greater than or equal to 70%, the corresponding strong convection suspicious target in the suspicious target area is displayed, otherwise it is not displayed, so as to obtain the strong convection occurrence prediction target area.

3. The tornado collaborative observation method based on detection equipment according to claim 2, characterized in that: Step S12 includes the following steps: Step S121: transmitting corresponding S-band electromagnetic waves by a new generation S-band weather radar to perform regional meteorological echo scanning on meteorological components behind the atmosphere in the suspicious target area to generate echo signals of meteorological components in the suspicious area; Step S122: analyzing the signal strength and reflectivity of the meteorological component echo signal in the suspicious area to obtain meteorological echo signal strength data and meteorological echo signal reflectivity data in the suspicious area; Step S123: performing a signal echo spatiotemporal distribution analysis of the meteorological components behind the atmosphere in the suspicious target area based on the suspicious area meteorological echo signal strength data and the suspicious area meteorological echo signal reflectivity data to generate a suspicious area meteorological signal echo spatiotemporal distribution map; Step S124: performing a meteorological spatiotemporal distribution field simulation analysis on the suspicious target area according to the spatiotemporal distribution map of meteorological signal echoes in the suspicious area to generate a spatiotemporal distribution field of meteorological echoes in the suspicious area; Step S125: Real-time monitoring of the spatiotemporal distribution of meteorological echoes in the suspicious area is performed to obtain meteorological echo data in the suspicious area, wherein the meteorological echo data in the suspicious area includes corresponding meteorological conditions and cloud density changes in the suspicious target area.

4. The tornado collaborative observation method based on detection equipment according to claim 2, characterized in that: The regional severe convection feature matching calculation formula in step S14 is specifically: ； Where, For suspicious target areas at time and location points The matching degree of strong convection characteristics in the corresponding suspicious area is: is the spatial domain of the suspicious target area, is the start time of the time range, is the end time of the time range, For suspicious target areas at time and location points The corresponding micro-pressure parameters of the suspicious area are: For suspicious target areas at time and location points The corresponding meteorological parameters of the suspicious area are as follows: is the characteristic parameter field of severe convection in the suspicious target area, is the weight coefficient of the strong convection characteristics, For suspicious target areas at time and location points The corresponding micro-pressure field gradient is: For suspicious target areas at time and location points The corresponding meteorological field gradient is is the gradient adjustment coefficient of the strong convection characteristic field, is an exponential function, is the reference position of the center of the suspicious target area, is the spatial position attenuation coefficient, is the correction coefficient for the matching degree of severe convective characteristics in the suspicious area.

5. The tornado collaborative observation method based on detection equipment according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: obtaining the solar radio radiation intensity at the corresponding signal wavelength of the corresponding radar detection equipment in the tornado detection network under the current solar radiation; Step S22: Calculate the received signal power of the corresponding radar detection equipment in the tornado detection network based on the solar radio radiation intensity at the corresponding signal wavelength under the current solar radiation to obtain the theoretical value of the radar detection received signal power. The calculation formula of the theoretical value of the radar detection received signal power is specifically: ； in, is the theoretical value of radar detection received signal power, is the radar receiving frequency bandwidth, is the radar antenna gain, is the radar signal wavelength, The wavelength of the radar signal The intensity of solar radio radiation at The radar receiving signal frequency; Step S23: obtaining a measured value of the radar detection received signal power, and performing a power difference calculation based on the measured value of the radar detection received signal power and a theoretical value of the radar detection received signal power to obtain a radar detection received signal power difference; Step S24: performing solar radio calibration processing on the corresponding radar detection equipment in the tornado detection network based on the difference in radar detection received signal power to generate a tornado detection calibration optimization network; Step S25: triggering a tornado detection strategy coordination instruction based on the corresponding strong convection suspicious target within the strong convection prediction target area, and applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network to perform three-dimensional coordinated observation of the strong convection target corresponding to the strong convection suspicious target within the strong convection prediction target area to generate a three-dimensional radar field of the strong convection target.

6. The tornado collaborative observation method based on detection equipment according to claim 5, characterized in that: Step S25 includes the following steps: Step S251: triggering the built-in policy control server to respond and send a tornado detection policy coordination instruction based on the existence of a suspected strong convection target corresponding to the strong convection occurrence prediction target area; Step S252: applying the tornado detection strategy coordination instruction to the tornado detection calibration optimization network and generating a tornado detection coordination scanning command in response; Step S253: applying the tornado detection collaborative scanning command to the corresponding search radar in the tornado detection calibration optimization network to perform target area detection scanning on the corresponding strong convection suspected targets within the strong convection occurrence prediction target area, so as to generate the target scanning signal echo intensity corresponding to each strong convection occurrence target area; Step S254: applying the tornado detection collaborative scanning command to the corresponding optical camera in the tornado detection calibration optimization network to perform target synchronous acquisition scanning on the corresponding strong convection suspected targets in the strong convection prediction target area to generate the target area size and target area position corresponding to each strong convection target area; calculating the scanning azimuth angle range based on the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convection target area to generate the strong convection target area scanning azimuth angle distribution range; Step S255: Apply the tornado detection collaborative scanning command to the corresponding ultra-fine radar in the tornado detection calibration optimization network and perform fixed-point collaborative fan scanning tracking and three-dimensional analysis on the corresponding strong convection suspected targets in the strong convection prediction target area based on the scanning azimuth distribution range of the strong convection target area to generate a three-dimensional radar field of the strong convection target.

7. The tornado collaborative observation method based on detection equipment according to claim 6, characterized in that: Calculating the scanning azimuth angle range according to the target scanning signal echo intensity, target area size, and target area position corresponding to each strong convection target area in step S254 includes the following steps: Performing spatial coordinate system conversion according to the target area position corresponding to each severe convection target area to generate a severe convection target area position spatial coordinate system; Based on the spatial coordinate system of the strong convection target area, the target area size corresponding to each strong convection target area is measured and calculated to obtain the regional scale characteristic index corresponding to each strong convection target area, where the regional scale characteristic index includes the regional major axis, regional minor axis and regional total area size; Based on the spatial coordinate system of the strong convection target area, the scanning azimuth angle range of the target scanning signal echo intensity and regional scale characteristic indicators corresponding to each strong convection target area are calculated to generate the scanning azimuth angle distribution range of the strong convection target area.

8. The tornado collaborative observation method based on detection equipment according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: obtaining a severe convective weather activity characteristic data matrix and an environmental condition distribution climate characteristic data matrix corresponding to the severe convective suspicious target in the severe convective target three-dimensional radar field; Step S32: Based on the severe convective weather activity characteristic data matrix and the environmental condition distribution climate characteristic data matrix, a canonical correlation analysis expression is used to perform a canonical correlation calculation on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field to obtain a canonical correlation coefficient between the strong convective weather activity corresponding to the strong convective suspicious targets and the environmental conditions; Among them, the canonical correlation analysis expression is specifically: ； Where, is the canonical correlation coefficient, is the coefficient vector corresponding to the typical variables in the characteristic data matrix of severe convective weather activities, is the coefficient vector corresponding to the typical variables in the climate characteristic data matrix of environmental condition distribution, is the vector transpose symbol, is the characteristic data matrix of severe convective weather activities, is the climate characteristic data matrix of environmental conditions distribution, is the expected value, is the covariance matrix between severe convective weather activities and environmental conditions, is the autocovariance matrix of severe convective weather activity data, is the autocovariance matrix of environmental condition data; Step S33: performing target equal-correlation classification on the corresponding strong convective suspicious targets in the three-dimensional radar field based on the typical correlation coefficient between the strong convective weather activities corresponding to the strong convective suspicious targets and the environmental conditions, so as to obtain a set of strong convective target equal-correlation; Step S34: performing target strong correlation region fusion division on the strong convective suspicious targets corresponding to the strong convective target three-dimensional radar field based on the strong convective target equal strong correlation set to generate a strong convective target equal strong correlation region radar field; Step S35: Gridding the radar field of the strong convective target and other strong correlation areas using the spatiotemporal adaptive reflectivity fusion model to generate a gridded strong convective grid field of the strong correlation areas; performing strong convective feature analysis on the gridded strong convective grid field of the strong correlation areas to generate a tornado and other strong convective feature dataset, wherein the tornado and other strong convective feature dataset includes horizontal polarization radar reflectivity. , horizontal and vertical dual polarization differential reflectivity , radar echo correlation coefficient , horizontal and vertical dual polarization rotation angle And the horizontal and vertical dual polarization differential phase standard deviation ,in is the differential phase measured by the radar under horizontal and vertical dual polarization.

9. The tornado collaborative observation method based on detection equipment according to claim 1, characterized in that: The expression corresponding to the spatiotemporal adaptive reflectivity fusion model in step S35 is specifically: ； Where, Output value of the spatiotemporal adaptive reflectivity fusion model, indicating the center position of the grid and time The gridded strong convective reflectivity value at is the maximum number of grid points, is the item index of the horizontal grid point, is the item index of the vertical grid point, Horizontal grid With vertical grid The weight coefficient matrix between For the The horizontal position parameters corresponding to the horizontal grid points, For the The vertical position parameters corresponding to the vertical grid points, At the grid point and time The original radar observed reflectivity value at At the grid point and the previous moment The original radar observed reflectivity value at At the grid point and the previous moment Gridded strong convective reflectivity values ​​at .

10. The tornado collaborative observation method based on detection equipment according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: obtaining a preset tornado feature library; Step S42: Construct a tornado and other strong convection identification model based on the tornado feature library and the tornado and other strong convection feature data set, and calculate the dual-polarization tornado vortex characteristic value based on the tornado and other strong convection identification model and tornado debris characteristic values ; Step S43: The dual polarized tornado vortex characteristic value and tornado debris characteristic values Substitute the tornado characteristic membership function corresponding to the severe convection identification model such as tornado to perform membership metric calculation to obtain the severe convection target tornado characteristic membership degree; Step S44: performing tornado identification analysis on the radar field of the corresponding strong convective target and other strongly correlated areas based on the tornado characteristic membership of the strong convective target to generate a grid point area prone to tornado occurrence.

Images