Automatic tropical cyclone tracking method based on relative threshold value

Abstract

The application relates to the field of tropical cyclone tracking, in particular to a tropical cyclone automatic tracking method based on a relative threshold value. The method comprises the following steps: obtaining meteorological historical data through a numerical weather prediction model, obtaining relative vorticity and warm core intensity of historical tropical cyclones according to the meteorological historical data; collecting historical tropical cyclone data, calculating relative thresholds of the relative vorticity and the warm core intensity according to the historical tropical cyclone data; obtaining meteorological data at different times according to the numerical weather prediction model, obtaining candidate areas at different times according to the relative thresholds and the meteorological data; obtaining multiple sets of longitude and latitude data and intensity data of tropical cyclone centers at different times according to the candidate areas at different times; and obtaining path and intensity information of each tropical cyclone changing over time by using a clustering algorithm according to the longitude and latitude data and the intensity data. The application solves the problems of low tracking efficiency and poor precision of multiple tropical cyclones in the prior art.

Description

Technical Field

[0001] This invention relates to the field of tropical cyclone tracking, specifically to an automatic tropical cyclone tracking method based on a relative threshold. Background Technology

[0002] Tropical cyclones are major meteorological disasters affecting many regions of the world. Their formation, development, and dissipation are complex processes, exhibiting different characteristics under different climatic conditions. Utilizing high-resolution numerical weather prediction models and meteorological data to locate and track tropical cyclones is one of the main fundamental tools for forecasting and studying tropical cyclones.

[0003] The main relevant content of existing technologies: The existing mainstream tropical cyclone tracking algorithms are divided into two parts. One part is tropical cyclone positioning: based on the main characteristics of tropical cyclones, the spatial position of tropical cyclones at different times is located by setting a fixed threshold. The other part is tracking: based on the average speed of tropical cyclone movement and the order of their positions, the positions of tropical cyclones within a certain range are connected to form the path of the tropical cyclone.

[0004] Problems and shortcomings of existing technologies: First, current mainstream algorithms lack completeness in tracking high-resolution weather forecast models (such as 3.5km and 1km spatial resolution), manifested in significant discrepancies between the number of tropical cyclones tracked and observed at different intensities. Most existing mainstream tracking algorithms were developed relatively early, based on coarse-resolution atmospheric numerical models, and did not consider the characteristics of tropical cyclones at high resolution or even ultra-high resolution. Therefore, serious omissions occur in the latest ultra-high resolution forecast models, resulting in a significantly lower number of tracked tropical cyclones compared to reality. Secondly, mainstream tropical cyclone tracking schemes are inefficient and time-consuming in ultra-high resolution weather forecast models. This is because mainstream tracking schemes track tropical cyclones one by one according to time and spatial location. In high-resolution models, when multiple potential tropical cyclone paths appear, mainstream tracking schemes require a long time. When processing multiple forecast results, mainstream tracking schemes are inefficient and require a large amount of runtime. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an automatic tropical cyclone tracking method based on relative thresholds, which solves the problems of low tracking efficiency and poor accuracy for multiple tropical cyclones in existing technologies.

[0006] To achieve the above objectives, the present invention provides an automatic tropical cyclone tracking method based on relative thresholds, comprising: acquiring historical meteorological data through a numerical weather prediction model, and obtaining the relative vorticity and warm core intensity of historical tropical cyclones based on the historical meteorological data; collecting historical tropical cyclone data, and calculating the relative thresholds for the relative vorticity and warm core intensity based on the historical tropical cyclone data; acquiring meteorological data at different times based on the numerical weather prediction model, and obtaining candidate regions at different times based on the relative thresholds and the meteorological data; obtaining latitude and longitude data and intensity data of multiple sets of tropical cyclone centers at different times based on the candidate regions at different times; and obtaining the path and intensity information of each tropical cyclone over time using a clustering algorithm based on the latitude and longitude data and the intensity data.

[0007] This invention acquires meteorological data through numerical weather prediction models and combines it with historical tropical cyclone data to comprehensively extract the value of the data. It calculates relative vorticity, warm core intensity, and relative thresholds using historical meteorological data, providing a basis for subsequent precise selection of candidate regions and fully leveraging the guiding role of the data. Clustering algorithms are used to derive the path and intensity information of tropical cyclones over time, which helps to accurately analyze the development dynamics of tropical cyclones and improves the overall accuracy of tracking tropical cyclone paths and intensity.

[0008] Optionally, obtaining the relative vorticity and warm core intensity of historical tropical cyclones based on the historical meteorological data includes: extracting the meridional wind component and zonal wind component at a distance of 10m above the ground from the historical meteorological data; calculating the relative vorticity of the tropical cyclone based on the meridional wind component and the zonal wind component; extracting the first average temperature and the second average temperature of the 200hPa isobaric surface and the 500hPa isobaric surface from the historical meteorological data, respectively; and calculating the warm core intensity of the tropical cyclone based on the first average temperature and the second average temperature.

[0009] This invention accurately acquires key meteorological elements by extracting meridional and zonal wind components at specific altitudes and different isobaric surface temperatures. When calculating relative vorticity, the meridional and zonal wind components effectively reflect the rotational characteristics of air near tropical cyclones, providing a basis for determining their dynamic structure. Furthermore, calculating the warm core intensity based on different isobaric surface temperatures measures the thermal characteristics of tropical cyclones. These calculation methods work together to comprehensively present the thermal and dynamic properties of tropical cyclones, laying a solid data foundation for subsequent calculations of relative thresholds, identification of candidate regions, and tracking of tropical cyclone paths and intensity changes, thereby improving the accuracy and reliability of tropical cyclone tracking.

[0010] Optionally, calculating the relative thresholds for the relative vorticity and the warm core intensity based on the historical tropical cyclone data includes: extracting historical relative vorticity values ​​and historical warm core structure values ​​within a 2° spherical neighborhood of the historical tropical cyclone center from the historical tropical cyclone data using statistical analysis based on the relative vorticity and the warm core intensity; calculating the historical relative vorticity maximum and historical warm core structure maximum for each historical tropical cyclone based on the historical relative vorticity values ​​and historical warm core structure values; and calculating the 95th percentiles of the historical relative vorticity maximum and historical warm core structure maximum to obtain the relative thresholds.

[0011] This invention focuses on selecting data from the 2° spherical neighborhood of historical tropical cyclone centers to accurately locate regions closely related to tropical cyclones, avoiding interference from other irrelevant regions and ensuring the effectiveness and relevance of the extracted data. This lays the foundation for accurately calculating relative thresholds. Statistical analysis is used to calculate historical relative vorticity maxima, historical warm core structure maxima, and their 95th quantiles as relative thresholds, effectively reflecting the distribution of tropical cyclone-related characteristics in historical data. This statistical approach based on a large amount of historical data minimizes the influence of random factors and improves the scientific rigor and accuracy of the thresholds.

[0012] Optionally, the relative threshold satisfies the following formula:

[0013]

[0014] This is a relative threshold. For the candidate values ​​of the variable, For indicator functions, For the first The historical relative vorticity maximum or historical warm core structure maximum of a historical tropical cyclone. The number of all historical tropical cyclones used to construct the empirical distribution.

[0015] The relative threshold formula of this invention is based on an empirical distribution constructed from a large amount of historical tropical cyclone data. It can fully explore historical information, reduce random errors, and make the threshold more stable and representative. Furthermore, it uses an indicator function to determine whether the conditions for candidate variable values ​​are met, and precisely defines the relative threshold in a rigorous mathematical form, thereby improving the calculation accuracy of the relative threshold.

[0016] Optionally, obtaining candidate regions for different time periods based on the relative threshold and the meteorological data includes: setting constraints using the relative threshold and filtering the meteorological data using the constraints to obtain candidate regions for different time periods;

[0017] The candidate regions satisfy the following formula:

[0018]

[0019] in, This represents the relative threshold for relative vorticity. Relative vorticity in meteorological data. The relative threshold for the intensity of warmth. The intensity of warmth in meteorological data.

[0020] This invention filters meteorological data from two key perspectives—dynamic and thermodynamic—by setting constraints in two dimensions: relative vorticity and warm core intensity, thus comprehensively and accurately capturing the relevant characteristics of tropical cyclones. This multi-condition joint screening method avoids the limitations of a single indicator and improves the accuracy and reliability of the screening.

[0021] Optionally, the intensity data includes the maximum 10m wind speed and the minimum sea level pressure. Obtaining multiple sets of latitude and longitude data and intensity data of tropical cyclone centers at different times based on the candidate regions at different times includes: calculating the 10m wind speed at different times using the meteorological data based on the candidate regions at different times; determining the vorticity maximum point within the candidate regions, and determining the wind speed minimum point at different times using the 10m wind speed within a 2° spherical neighborhood centered on the vorticity maximum point; obtaining the latitude and longitude data of the tropical cyclone centers at different times based on the wind speed minimum points at different times; and calculating the maximum 10m wind speed and the minimum sea level pressure within a 2° spherical neighborhood of the tropical cyclone centers at different times using the meteorological data.

[0022] This invention calculates 10m wind speeds based on candidate regions and uses the vorticity maxima to determine the wind speed minima, thus accurately pinpointing the center's latitude and longitude. It fully considers the rotational characteristics of tropical cyclones, effectively eliminating interference and improving positioning accuracy. Regarding intensity data acquisition, it calculates the maximum 10m wind speed and minimum sea-level pressure within a specific spherical neighborhood around the determined center, ensuring that the obtained data accurately reflects the core intensity of the tropical cyclone.

[0023] Optionally, obtaining the path and intensity information of each tropical cyclone over time using a clustering algorithm based on the latitude and longitude data and the intensity data includes: constructing a three-dimensional feature vector for each candidate region based on the latitude and longitude data and the intensity data; constructing an adjacency structure for the candidate regions based on the three-dimensional feature vectors; classifying the central nodes of the candidate regions based on the adjacency structure; constructing a density connectivity cluster based on the classification results; and obtaining the path and intensity information of each tropical cyclone over time based on the density connectivity cluster.

[0024] This invention constructs a three-dimensional feature vector using latitude, longitude, and intensity data, effectively integrating key information about tropical cyclones and providing a structured data foundation for subsequent analysis. The adjacency structure built based on the three-dimensional feature vector accurately measures the correlation between candidate regions, considering spatial location and intensity differences, thus conforming to the actual characteristics of tropical cyclones. Classifying central nodes and constructing density connectivity clusters enables effective identification and segmentation of tropical cyclones. This clustering method fully explores hidden connections between data, accurately improving the tracking accuracy of tropical cyclones.

[0025] Optionally, constructing the adjacency structure of the candidate region based on the three-dimensional feature vector includes: calculating the spatial distance and intensity difference between two tropical cyclone centers at different times based on the three-dimensional feature vector; constructing a comprehensive distance metric function between the two tropical cyclone centers at different times based on the spatial distance and the intensity difference; and constructing the adjacency structure of the candidate region based on the comprehensive distance metric function.

[0026] This invention comprehensively considers the spatiotemporal and intensity variations of tropical cyclones by calculating the spatial distances and intensity differences between their centers at different times. Based on this, a comprehensive distance metric function is constructed that organically combines spatial and intensity factors, more realistically reflecting the similarity and correlation between subtropical cyclones at different times. The resulting adjacency structure can accurately delineate the relationships between tropical cyclones at different times, improving the scientific rigor and accuracy of the adjacency structure construction.

[0027] Optionally, constructing the adjacency structure of the candidate region based on the comprehensive distance metric function includes: acquiring historical path data of tropical cyclones, and obtaining historical spatial data and historical intensity data of two historical tropical cyclone centers at different times based on the historical path data; calculating the 95th percentile of the historical spatial data and the historical intensity data respectively to obtain historical spatial thresholds and historical intensity thresholds; calculating the cluster radius based on the historical spatial thresholds and the historical intensity thresholds; and constructing the adjacency structure of the candidate region based on the cluster radius and the comprehensive distance metric function.

[0028] Calculating spatial and intensity thresholds using historical path data allows for the full extraction of the value of historical data, as these thresholds represent the general characteristics of tropical cyclones in terms of spatial and intensity variations. The clustering radii calculated using these spatial and intensity thresholds provide a scientific standard for defining the associations between subtropical cyclone centers at different times. Combining this with a comprehensive distance metric function to construct an adjacency structure further enhances the scientific rigor and accuracy of the adjacency structure construction.

[0029] The adjacency structure satisfies the following formula:

[0030]

[0031] in, To be based on three-dimensional feature vectors The constructed adjacency structure, To be related to three-dimensional feature vectors The nearest neighbor and the combined distance metric function is less than the cluster radius value. Other three-dimensional feature vectors, For calculation using the spherical cosine formula and Spatial distance between them For intensity weight, For calculation and The intensity difference value, This represents the cluster radius value.

[0032] The adjacency structure formula of this invention comprehensively considers spatial distance and intensity difference, uses the spherical cosine formula to calculate the spatial distance value, and combines intensity weight to measure the intensity difference value. It comprehensively and accurately describes the relationship between three-dimensional feature vectors. The formula sets the cluster radius value as a judgment criterion, which can effectively filter out other vectors that are close to the target three-dimensional feature vector and meet the requirements of the comprehensive distance measurement function, thus constructing a more reasonable adjacency structure and improving the accuracy of adjacency structure calculation. Attached Figure Description

[0033] Figure 1 This is a flowchart of the automatic tropical cyclone tracking method based on relative thresholds according to an embodiment of the present invention. Detailed Implementation

[0034] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0035] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0036] Please see Figure 1 , Figure 1 This is a flowchart of the automatic tropical cyclone tracking method based on relative thresholds according to an embodiment of the present invention, as shown below. Figure 1 The method shown includes the following steps:

[0037] Step S1: Obtain historical meteorological data through a numerical weather prediction model, and obtain the relative vorticity and warm core intensity of historical tropical cyclones based on the historical meteorological data.

[0038] The specific steps for obtaining the relative vorticity and warm core intensity of historical tropical cyclones based on the historical meteorological data include the following:

[0039] Step S101: Extract the meridional wind component and zonal wind component at a distance of 10m above the ground from the historical meteorological data.

[0040] In this embodiment, numerical weather prediction (NMR) models are a scientific method for making weather forecasts based on mathematical models of the atmosphere and using current weather conditions as input data. Based on the fundamental laws of fluid mechanics and thermodynamics, it describes atmospheric motion using a set of mathematical formulas, which are then numerically solved using a high-performance computer. Before computation, a large amount of current meteorological observation data needs to be collected, such as temperature, air pressure, humidity, precipitation, and wind speed obtained from weather stations, meteorological satellites, and weather balloons, as initial conditions for the model. Numerical weather prediction models are divided into global models and regional models. Global models aim to accurately predict global weather; regional models focus on weather conditions in specific regions. In practical applications, multiple models are often used in combination to reduce the uncertainty of weather forecasts and extend the predictable timeframe.

[0041] First, relying on advanced numerical weather prediction models and high-performance computing clusters, and using massive amounts of meteorological observation data as the initial field, complex fluid dynamics and thermodynamic equations are used to generate historical meteorological data covering many years and a global scale. Then, using the professional meteorological data processing software GrADS (Grid Analysis and Display System), specific control statements and script code are written in the GrADS interface to call its rich library of data reading and analysis functions. Based on precise altitude commands (10m above the ground), meticulous filtering is performed from the vast historical meteorological dataset to extract the two key elements: meridional wind components and zonal wind components.

[0042] Wind field data at a height of 10m above the ground is closely related to surface friction and has a crucial impact on the formation and maintenance of the near-surface circulation structure of tropical cyclones. The meridional wind component represents the north-south air transport, while the zonal wind component reflects the east-west air flow changes. Their synergistic effect is significant for characterizing the rotation, movement, and energy exchange of tropical cyclones, and also lays a solid data foundation for subsequent in-depth calculations of important parameters such as relative vorticity.

[0043] Step S102: Calculate the relative vorticity of the tropical cyclone based on the meridional wind component and the zonal wind component.

[0044] In this embodiment, relative vorticity is a physical quantity that measures the degree of rotation of air particles around a vertical axis. For strongly rotating weather systems such as tropical cyclones, calculating their relative vorticity can help to gain a deeper understanding of their dynamic characteristics and development mechanisms.

[0045] First, the study area is meticulously divided into small grids, each corresponding to specific meridional and zonal wind component values. Then, the zonal wind components of adjacent grid points on either side of a given grid point are analyzed to identify their differences. Combined with grid distance information, the east-west variation of the zonal wind component is obtained. Similarly, the meridional wind components of adjacent grid points on either side of the same grid point are analyzed to identify their differences. Combined with grid distance information, the north-south variation of the meridional wind component is obtained. Finally, the relative vorticity of the tropical cyclone is calculated by combining the wind component variations in these two directions.

[0046] The relative vorticity satisfies the following formula:

[0047]

[0048] in, For longitude ,latitude and time Relative vorticity below, The meridional wind component is located 10 meters above the ground. This represents the zonal wind component at an altitude of 10 meters above the ground. The coordinates represent the distance in the east-west direction. These are the distance coordinates in the north-south direction.

[0049] Step S103: Extract the first average temperature and the second average temperature of the 200hPa isobaric surface and the 500hPa isobaric surface from the historical meteorological data, respectively.

[0050] In this embodiment, GrADS software is used to accurately locate the 200 hPa and 500 hPa isobaric surfaces according to standard atmospheric pressure layer settings. The software's temperature data filtering function is then used to select the corresponding isobaric surface temperature data from a vast amount of historical meteorological data. These temperature data are discretely distributed across different spatial locations. Next, for each point where the average temperature is to be calculated, all temperature data points are collected within a circular neighborhood with a radius of 2° centered on that point. Then, a suitable mathematical statistical method, such as the arithmetic mean, is used to process the temperature values ​​of these data points to calculate the spatial average temperature within that neighborhood. After completing the above operations for all points on the 200 hPa isobaric surface, the resulting spatial average temperatures are averaged again to obtain the average temperature of the 200 hPa isobaric surface, which is the first average temperature. Similarly, the average temperature of the 500 hPa isobaric surface is the second average temperature.

[0051] Step S104: Calculate the warm core intensity of the tropical cyclone based on the first average temperature and the second average temperature.

[0052] In this embodiment, warm core intensity is an indicator that measures the degree to which the temperature in the central region of a tropical cyclone is higher than that in the surrounding environment. Its value is usually obtained by calculating the temperature difference between the system center and a certain range of the periphery, or the temperature difference between different isobaric surfaces; the larger the difference, the stronger the warm core intensity.

[0053] The warmth intensity satisfies the following formula:

[0054]

[0055] in, For longitude ,latitude and time The intensity of warmth, The first average temperature, This is the second average temperature.

[0056] Step S2: Collect historical tropical cyclone data, and calculate the relative thresholds of the relative vorticity and the warm core intensity based on the historical tropical cyclone data.

[0057] The calculation of the relative thresholds for relative vorticity and warm core intensity based on the historical tropical cyclone data specifically includes the following sub-steps:

[0058] Step S201: Based on the relative vorticity and the warm core intensity, the historical relative vorticity value and historical warm core structure value within the 2° spherical neighborhood of the historical tropical cyclone center are extracted from the historical tropical cyclone data using statistical analysis.

[0059] In this embodiment, historical tropical cyclone data is comprehensively collected through professional meteorological data platforms, such as the International Best Tracks Archive (IBTrACS) of the World Meteorological Organization, the official database of the National Hurricane Center (NHC) of the United States, and relevant data resource databases of the China Meteorological Administration. This data contains rich information such as the formation time, movement trajectory, center latitude and longitude, and intensity changes of tropical cyclones. It is also important to understand that the relative vorticity and warm core intensity data calculated from historical meteorological data are global data calculated based on meteorological elements at different grid points. Based on the center latitude and longitude of each tropical cyclone, a 2° spherical neighborhood is determined using geographic information technology and related algorithms, with each tropical cyclone center as the center. Within this neighborhood, the corresponding relative vorticity and warm core intensity data are selected. Then, statistical analysis methods, such as calculating the maximum, average, and median, are used to process the relative vorticity data within the neighborhood to obtain historical relative vorticity values ​​that represent the characteristics of that neighborhood. Similarly, similar statistical processing is performed on the warm core intensity data within the neighborhood to obtain historical warm core structure values.

[0060] Step S202: Calculate the historical relative vorticity maximum and historical warm core structure maximum for each historical tropical cyclone based on the historical relative vorticity value and the historical warm core structure value.

[0061] In this embodiment, for each historical tropical cyclone, a 2° spherical neighborhood is first defined based on its center's latitude and longitude. Then, within this neighborhood, the relative vorticity values ​​at different locations at a specific time are examined one by one. By comparing these values, the largest value is identified; this value is the historical relative vorticity maximum of the historical tropical cyclone at that moment within the 2° spherical neighborhood. Similarly, the warm core structure values ​​at different locations within the neighborhood at a specific time are comprehensively compared, and the largest value is selected; this is the historical warm core structure maximum of the historical tropical cyclone at that moment within the 2° spherical neighborhood.

[0062] The historical relative vorticity maxima and the historical warm core structure maxima satisfy the following formula:

[0063]

[0064] in, This represents the historical maximum relative vorticity or the historical maximum warm core structure. To obtain the longitude ,latitude and time The maximum value function of historical relative vorticity or historical warm core structure within a spherical neighborhood no more than 2° away from the center of the tropical cyclone.

[0065] Step S203: Calculate the 95th percentile of the historical relative vorticity maximum and the historical warm core structure maximum to obtain the relative threshold.

[0066] In this embodiment, the number of historical tropical cyclones involved in the calculation is first determined, along with the historical relative vorticity maxima or historical warm core structure maxima corresponding to each historical tropical cyclone. Next, within a reasonable range, the value of variable x is continuously selected. For each selected x, all relative vorticity maxima or historical warm core structure maxima are examined one by one. It is determined whether the relative vorticity maxima or historical warm core structure maxima is less than or equal to x; if so, it is recorded as 1, otherwise as 0. These 1s and 0s are accumulated and divided by N to calculate the proportion. This process is continued until an x ​​that makes the proportion reach or exceed 95% is found. Among all x values ​​that meet the requirements, the largest lower bound is found; this largest lower bound is the relative threshold we need. By performing the above operations on the historical relative vorticity maxima and historical warm core structure maxima respectively, their corresponding relative thresholds can be obtained.

[0067] The 95th percentile is a common quantile indicator, representing that 95% of the data values ​​in a dataset are less than or equal to that quantile, while only 5% of the data values ​​are greater than that quantile. It reflects the locational characteristics of the data distribution and is often used to exclude outliers and determine the "prevailing level" or "critical value" of the data.

[0068] The relative threshold satisfies the following formula:

[0069]

[0070] This is a relative threshold. For the candidate values ​​of the variable, For indicator functions, For the first The historical relative vorticity maximum or historical warm core structure maximum of a historical tropical cyclone. The number of all historical tropical cyclones used to construct the empirical distribution.

[0071] Step S3: Obtain meteorological data at different times based on the numerical weather prediction model, and obtain candidate areas at different times based on the relative threshold and the meteorological data.

[0072] The specific steps for obtaining candidate regions at different times based on the relative threshold and the meteorological data include the following:

[0073] The relative threshold is used to set constraints, and the meteorological data is filtered using the constraints to obtain candidate regions at different times.

[0074] In this embodiment, relative thresholds for relative vorticity and warm core intensity are used as screening criteria to check the relative vorticity in the meteorological data. If the relative vorticity is not lower than the relative threshold, this part of the data is retained. Then, the warm core intensity is checked; if it is not lower than the relative threshold, it is also retained. Only areas that meet both conditions are selected. This process is repeated for meteorological data from different times to obtain areas that meet the requirements at different times, i.e., candidate areas for different time periods.

[0075] The candidate regions satisfy the following formula:

[0076]

[0077] in, This represents the relative threshold for relative vorticity. Relative vorticity in meteorological data. The relative threshold for the intensity of warmth. The intensity of warmth in meteorological data.

[0078] Step S4: Based on the candidate regions at different times, obtain multiple sets of latitude and longitude data and intensity data of tropical cyclone centers at different times.

[0079] The process of obtaining multiple sets of latitude, longitude, and intensity data of tropical cyclone centers at different times based on the candidate regions at different times specifically includes the following sub-steps:

[0080] Step S401: Calculate the 10m wind speed at different times using the meteorological data based on the candidate areas at different times.

[0081] In this embodiment, a 10m wind speed refers to the horizontal air velocity at a height of 10 meters above the ground. For each candidate region at any given time, relevant meteorological data is collected, including various factors such as air pressure, temperature, humidity, and wind direction. Then, specialized meteorological algorithms and models, such as numerical models based on atmospheric dynamics principles, are used to calculate the wind speed, taking into account the interrelationships between various meteorological elements within the candidate region. During the calculation process, the influence of horizontal and vertical atmospheric motion on wind speed, as well as the effects of topography and underlying surface factors, are fully considered. Through comprehensive analysis and calculation of this meteorological data, the 10m wind speed values ​​for the candidate regions at different times are finally obtained.

[0082] Step S402: Determine the vorticity maximum point within the candidate region, and use the 10m wind speed as the center to determine the wind speed minimum point at different times within a 2° spherical neighborhood.

[0083] In this embodiment, for each candidate region at each time point, the meteorological data within it is carefully analyzed, with a particular focus on vorticity, an important meteorological element. The vorticity value of each location point within the candidate region is precisely acquired, and a comprehensive and detailed analysis is conducted by comparing the magnitude of the vorticity values ​​point by point. Finally, the point with the largest vorticity value is identified and determined as the vorticity maximum point.

[0084] After successfully identifying the point of maximum vorticity, a 2° spherical neighborhood was accurately delineated using this point as a precise central reference, based on geospatial calculation methods. Within this specific neighborhood, corresponding 10m wind speed data were systematically collected for different time periods. Each time period's 10m wind speed data included wind speed information for various locations within the neighborhood.

[0085] Then, a detailed comparative analysis was conducted on the 10m wind speed values ​​at various locations within the neighborhood for each time period. During this process, through rigorous comparison and selection, the point with the lowest wind speed value at that time period was identified, and this point was designated as the wind speed minimum point for that time period. Following this procedure, corresponding wind speed minimum points were determined within the 2° spherical neighborhood for different time periods. The determination of these wind speed minimum points at different time periods provides crucial and precisely located foundational data for subsequent in-depth research into the structural characteristics and variation patterns of tropical cyclones at different times.

[0086] Step S403: Obtain the latitude and longitude data of the tropical cyclone center at different times based on the minimum wind speed points at different times.

[0087] The process of obtaining the latitude and longitude data of the tropical cyclone center at different times based on the minimum wind speed points at different times includes:

[0088] Extract the distribution characteristics of meteorological elements from the meteorological data;

[0089] Based on the distribution characteristics, the minimum wind speed is corrected to obtain the corrected minimum wind speed at different times;

[0090] Based on the minimum corrected wind speed values ​​at different times, the latitude and longitude data of the tropical cyclone center at different times can be obtained.

[0091] In this embodiment, the acquired meteorological data is comprehensively and meticulously analyzed to extract various meteorological elements, such as 10m wind speed, relative vorticity, warm core intensity, air pressure, temperature, humidity, and wind direction, and their distribution characteristics at different times and spatial locations. This includes the locations of high and low value areas for each element, their trends, gradient directions, and intensities. These characteristics are clearly presented through methods such as plotting contour maps and analyzing statistical data. Next, based on the extracted meteorological element distribution characteristics, the identified minimum wind speed points at different times are thoroughly evaluated. Considering that minimum wind speed values ​​may be affected by surrounding pressure gradients, topography, and the synergistic influence of other meteorological elements, if the minimum wind speed point is located in an area of ​​abnormal pressure gradient, or in complex terrain (such as valleys or leeward slopes of mountains), or when there is a significant contradiction with the distribution of other meteorological elements, appropriate adjustments are made based on relevant meteorological principles and empirical models to obtain corrected minimum wind speed values ​​at different times that more accurately reflect the actual situation. Finally, based on these modified minimum wind speed values ​​at different times, and combined with geographic information system technology and the structural characteristics and formation and development patterns of tropical cyclones, the spatial positional relationship of the modified minimum wind speed points and their correlation with the distribution of other meteorological elements are comprehensively analyzed. Their coordinates are then fine-tuned to determine the relatively accurate latitude and longitude data of the tropical cyclone center at different times.

[0092] In one alternative embodiment, the corrected minimum wind speed satisfies the following formula:

[0093]

[0094] in, For the first The corrected minimum wind speed for each time period. For the first The minimum wind speed at that time, and All are weights. The relative vorticity characteristic value is extracted from the distribution characteristics of meteorological elements. Pressure gradient feature values ​​extracted from the distribution characteristics of meteorological elements.

[0095] The above and The correlation between historical tropical cyclone data and the corresponding wind speed, vorticity, and pressure gradient can be used to confirm and calibrate the data.

[0096] Step S404: Calculate the maximum 10m wind speed and minimum sea level pressure within a 2° spherical neighborhood of the center of the tropical cyclone at different times using the meteorological data.

[0097] In this embodiment, the intensity data includes the maximum 10m wind speed and the minimum sea level air pressure.

[0098] Based on the known latitude and longitude data of subtropical cyclone centers at different times, a 2° spherical neighborhood is delineated for each time period, centered on the cyclone center, using Geographic Information System (GIS) technology. Using the professional meteorological data processing software GrADS, corresponding data subsets are precisely extracted from the acquired meteorological data according to the delineated area and time period. The focus is on extracting 10m wind speed and sea-level pressure data for each grid point within this neighborhood, while also extracting auxiliary meteorological elements such as temperature, humidity, and wind direction for subsequent comprehensive analysis. The extracted data is then systematically organized according to the latitude and longitude coordinates of the grid points, constructing a two-dimensional array structure that facilitates calculation and analysis. This ensures that each data point accurately corresponds to its geographical location, providing a clear data foundation for subsequent calculations of the maximum 10m wind speed and minimum sea-level pressure.

[0099] The 10m wind speed data within the neighborhood of each time period were processed using programming languages ​​such as Python. Leveraging the powerful array operations of the NumPy library, the wind speed data was converted into a NumPy array. By calling the array's maximum value function `np.max()`, each wind speed value in the array was quickly traversed and compared to determine the maximum 10m wind speed value within the 2° spherical neighborhood for that time period.

[0100] For sea level pressure data, a similar process to that used for calculating the maximum 10m wind speed is employed. The sea level pressure data is also organized into a NumPy array, and the minimum value in the array is obtained using the `np.min()` function, representing the lowest sea level pressure within a 2° spherical neighborhood of the subtropical cyclone center at that time. Since sea level pressure can be affected by factors such as topography and tides, the pressure data is corrected before calculation using high-precision topographic data and tidal models. For topographic factors, pressure measurements at different altitudes are corrected based on the formula for pressure variation with altitude; for tidal influences, pressure data for coastal areas are adjusted accordingly based on tidal forecast data. These correction steps eliminate interference from external factors, ensuring the accuracy and reliability of the calculated minimum sea level pressure.

[0101] Step S5: Based on the latitude and longitude data and the intensity data, a clustering algorithm is used to obtain the path and intensity information of each tropical cyclone over time.

[0102] The process of obtaining the time-varying path and intensity information of each tropical cyclone using a clustering algorithm based on the latitude and longitude data and the intensity data specifically includes the following sub-steps:

[0103] Step S501: Construct a three-dimensional feature vector for each candidate region based on the latitude and longitude data and the intensity data.

[0104] In this embodiment, latitude and longitude data include longitude and latitude data, and intensity data includes maximum 10m wind speed and minimum sea level pressure. To ensure comparability and computability of these different types and magnitudes of data, the latitude and longitude data, maximum 10m wind speed, and minimum sea level pressure are standardized. For latitude and longitude, a normalization mapping is performed based on the latitude and longitude span of the study area. For maximum 10m wind speed and minimum sea level pressure, the Z-score standardization method is used to eliminate the influence of dimensions. After standardization, the standardized longitude value is used as the first component of the three-dimensional feature vector, and the standardized latitude value is used as the second component. The standardized maximum 10m wind speed and minimum sea level pressure are linearly combined according to a certain weight (for example, the weights are determined based on historical data correlation analysis to make their contributions to the feature vector more reasonable), and the resulting value is used as the third component, thereby constructing a unique three-dimensional feature vector for each candidate region.

[0105] Step S502: Construct the adjacency structure of the candidate region based on the three-dimensional feature vector.

[0106] The construction of the adjacency structure of the candidate region based on the three-dimensional feature vector specifically includes the following sub-steps:

[0107] Step S50201: Calculate the spatial distance and intensity difference between the centers of two tropical cyclones at different times based on the three-dimensional feature vector.

[0108] In this embodiment, calculations are performed on vectors from any two different time periods. The first two dimensions of the three-dimensional feature vector represent the latitude and longitude information of the tropical cyclone center. Using the spherical cosine formula, the latitude and longitude values ​​corresponding to the two time periods are input to accurately calculate their spatial distance on the Earth's surface. This method considers the curvature of the Earth and can effectively reflect the actual spatial interval. The third dimension of the vector integrates the maximum 10m wind speed and the minimum sea level pressure information. Subtracting the values ​​of this dimension from the two time period vectors and taking the absolute value yields a preliminary value of the intensity difference. Considering that wind speed and air pressure have different degrees of influence on intensity, their respective weights are determined based on historical data and meteorological principles. The preliminary intensity difference value is then weighted and corrected to obtain a value that more accurately reflects the intensity difference between the centers of the two tropical cyclones at different time periods.

[0109] In an optional embodiment, the spatial distance satisfies the following formula:

[0110]

[0111] in, The three-dimensional eigenvectors calculated using the spherical cosine formula. sum and three-dimensional eigenvectors The nearest neighbor and the combined distance metric function is less than the cluster radius value. Other three-dimensional feature vectors Spatial distance between them The average radius of the Earth and These are the longitudes of the centers of two tropical cyclones at different times. and These are the dimensions of the centers of two tropical cyclones at different times.

[0112] In an optional embodiment, the intensity difference satisfies the following formula:

[0113]

[0114] in, for and The intensity difference value, and These are the intensity indices of the centers of two tropical cyclones at different times.

[0115] Step S50202: Construct a comprehensive distance metric function between the centers of two tropical cyclones at different times based on the spatial distance and the intensity difference.

[0116] In this embodiment, the comprehensive distance metric function is a mathematical tool that combines spatial distance and intensity differences. Through a specific calculation method, it yields a numerical value that comprehensively reflects the degree of difference between the three-dimensional feature vectors corresponding to the subtropical cyclone centers at different times. This value can be used to determine the similarity or correlation between objects, and in tropical cyclone research, it can assist in analyzing the changes in subtropical cyclone centers at different times.

[0117] To accurately measure the comprehensive differences between subtropical cyclone centers at different times, a comprehensive distance metric function is constructed when studying them. First, two three-dimensional eigenvectors are determined, and the spatial distance between them is calculated using the spherical cosine formula, reflecting their spatial location differences. Then, the intensity difference between the two eigenvectors is calculated, showing their differences in intensity. Weights are then assigned to reflect the importance of the intensity difference in the comprehensive measurement. Finally, the spatial distance and intensity difference values ​​are squared respectively, combined according to the weights, and the square root is taken to obtain the comprehensive distance metric value, thus quantitatively describing the comprehensive spatial and intensity differences between two subtropical cyclone centers at different times.

[0118] The comprehensive distance metric function satisfies the following formula:

[0119]

[0120] in, For the comprehensive distance metric function, It is a three-dimensional feature vector. To be related to three-dimensional feature vectors The nearest neighbor and the combined distance metric function is less than the cluster radius value. Other three-dimensional feature vectors, For calculation using the spherical cosine formula and Spatial distance between them For calculation and The intensity difference value, For intensity weights.

[0121]

[0122] in, The standard deviation is the historical intensity difference.

[0123] Step S50203: Construct the adjacency structure of the candidate region according to the comprehensive distance metric function.

[0124] The construction of the adjacency structure of the candidate region based on the comprehensive distance metric function specifically includes the following sub-steps:

[0125] Step S5020301: Obtain historical track data of tropical cyclones, and obtain historical spatial data and historical intensity data of the centers of two different historical tropical cyclones based on the historical track data.

[0126] In this embodiment, these data record in detail the center latitude and longitude coordinates (spatial location) of tropical cyclones at different time points, along with the corresponding maximum 10m wind speed and minimum sea-level pressure (intensity indicators). For the two different time periods of the research target, the records for these two time periods are first located in the historical path data. The corresponding center latitude and longitude are extracted, and the spherical distance between them is calculated using the spherical cosine formula, serving as historical spatial data. Simultaneously, the maximum wind speed and minimum sea-level pressure for the corresponding time periods are extracted, and the intensity difference is calculated, serving as historical intensity data. During this process, the accuracy and completeness of the data must be carefully verified. Any missing or erroneous data is processed through data interpolation and cross-validation with other reliable data sources to ensure that the obtained spatial and intensity data accurately reflect the characteristics of the historical tropical cyclone centers at the two different time periods.

[0127] Step S5020302: Calculate the 95th percentile of the historical spatial data and the historical intensity data respectively to obtain the historical spatial threshold and the historical intensity threshold.

[0128] Collect spherical distance data between the centers of all subtropical cyclones at different times, and carefully arrange these data from smallest to largest to form an ordered data column. Next, count the total number of such data points, assuming there are n in total. Then, calculate 95% of n, which is n multiplied by 0.95, to obtain a value. If this value is an integer, find the corresponding data point in the ordered data column; this data point is the 95th quantile of the historical spatial data, i.e., the historical spatial threshold. If the value is not an integer, take the nearest integer smaller than it, let's say this integer is m. First, find the m-th data point in the data column, then find the (m+1)-th data point, and subtract the m-th data point from the (m+1)-th data point to obtain a difference. Subtract m from the non-integer value obtained earlier to obtain a decimal. Finally, add the product of this decimal and the difference to the m-th data point; the result is the historical spatial threshold.

[0129] Perform the same process on historical intensity data. Collect data reflecting changes in tropical cyclone intensity, such as maximum wind speed difference and minimum pressure difference, and arrange them in ascending order. Using the same method as for historical spatial data, calculate the 95th percentile of the data, and determine the 95th percentile based on whether the result is an integer. This percentile is the historical intensity threshold representing a boundary value for changes in tropical cyclone intensity.

[0130] Step S5020303: Calculate the cluster radius based on the historical spatial threshold and the historical intensity threshold.

[0131] In this embodiment, the cluster radius is calculated based on in-depth analysis of the historical characteristics of tropical cyclones, aiming to provide a reasonable quantitative standard for determining the correlation between subtropical cyclone centers at different times. The historical spatial threshold reflects the general upper limit of the spatial distance between adjacent sub-centers during the historical movement of a tropical cyclone; the distance between adjacent sub-centers of a tropical cyclone will not exceed the historical spatial threshold. Similarly, the historical intensity threshold represents a common limit for the intensity changes of tropical cyclones; their intensity changes will not exceed the historical intensity threshold.

[0132] When calculating the cluster radius, historical spatial thresholds and historical intensity thresholds are considered together. Since the spatial movement and intensity changes of tropical cyclones are crucial for determining their identity and development path, a specific calculation method combines the two. Typically, an intensity weight is determined based on historical data and meteorological principles; this weight reflects the importance of intensity changes in the overall assessment. The cluster radius is obtained by adding the square of the historical spatial threshold to the square of the product of the intensity weight and the historical intensity threshold, and then taking the square root of the sum.

[0133] The cluster radius plays a crucial role in subsequent tropical cyclone tracking analysis. When constructing the adjacency structure of candidate regions, it serves as a metric to determine whether the combined distance between the three-dimensional feature vectors of subtropical cyclone centers at different times is within a reasonable range. If the combined distance between two times is less than the cluster radius, they are considered to have a strong correlation and are likely to belong to the same tropical cyclone at different times, thus laying the foundation for accurately depicting the path and intensity information of tropical cyclones over time.

[0134] The cluster radius satisfies the following formula:

[0135]

[0136] in, The cluster radius is 1. For historical spatial threshold, For intensity weight, This is the historical intensity threshold.

[0137] Step S5020304: Construct the adjacency structure of the candidate region based on the clustering radius and the comprehensive distance metric function.

[0138] The adjacency structure satisfies the following formula:

[0139]

[0140] in, To be based on three-dimensional feature vectors The constructed adjacency structure, To be related to three-dimensional feature vectors The nearest neighbor and the combined distance metric function is less than the cluster radius value. Other three-dimensional feature vectors, For calculation using the spherical cosine formula and Spatial distance between them For intensity weight, For calculation and The intensity difference value, This represents the cluster radius value.

[0141] Step S503: Classify the center nodes of the candidate regions according to the adjacency structure.

[0142] In this embodiment, the adjacency structure defines the relationship between three-dimensional feature vectors. By combining a distance metric function and a clustering radius, it determines which nodes are adjacent to each other and may belong to the same tropical cyclone. Based on this, when classifying the central nodes, all central nodes in the candidate regions are traversed. For each node, other nodes adjacent to it with a combined distance smaller than the clustering radius are identified based on the adjacency structure, and these adjacent nodes are grouped into the same category. Specific algorithms and logic are used in the classification process. For example, starting with an unclassified node, it is marked as the starting node of a new category. Then, according to the adjacency relationship, all associated neighboring nodes are included in that category and marked as classified. The process continues to find the next unclassified node and repeats until all nodes are classified. Through this classification method, tropical cyclone central nodes with similar spatiotemporal and intensity characteristics can be grouped together, clearly distinguishing different individual tropical cyclones. This not only helps in analyzing the independent development path and intensity changes of each tropical cyclone over time but also avoids incorrect association of nodes from different cyclones, thereby greatly improving the accuracy and reliability of tropical cyclone tracking.

[0143] Step S504: Construct density connectivity clusters based on the classification results, and obtain the path and intensity information of each tropical cyclone over time based on the density connectivity clusters.

[0144] In this embodiment, density connectivity clustering is a method for clustering based on the density relationship between data points. In this invention, the central node in each category can be regarded as a set of data points with similar spatial and intensity characteristics. For these sets, the core points are first determined, that is, those nodes that contain a sufficient number of neighboring points in their neighborhood (the proximity relationship is determined by the adjacency structure and the cluster radius). The core points represent the relatively stable and significant state of the tropical cyclone at a certain moment.

[0145] Starting from the core point, nodes directly or indirectly connected to it are linked through density reachability relationships. If a node can reach another node through a series of density-connected core points, then the two nodes belong to the same density connectivity cluster. During the construction process, the cluster scope is continuously expanded, gradually incorporating nodes that meet the criteria to form complete clusters. Each cluster corresponds to a set of states of a tropical cyclone at different times.

[0146] This clustering method allows for the clear differentiation of individual tropical cyclones. For each cluster, the corresponding latitude and longitude data are extracted sequentially according to the time order of the nodes within the cluster. Connecting these latitude and longitude data in chronological order forms the movement path of the tropical cyclone. Simultaneously, based on the intensity data within the nodes, including the maximum 10m wind speed and minimum sea-level pressure, the intensity changes at different times can be visually presented. This process not only effectively integrates the various information obtained in previous steps but also delves into the intrinsic relationships between the data, providing solid data support for a comprehensive and accurate study of the development and evolution patterns of tropical cyclones.

[0147] The tropical cyclone is automatically tracked based on the path and intensity information.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. An automatic tropical cyclone tracking method based on relative thresholds, characterized in that, The method includes: Historical meteorological data are obtained through numerical weather prediction models, and the relative vorticity and warm core intensity of historical tropical cyclones are obtained based on the historical meteorological data. Collecting historical tropical cyclone data and calculating the relative thresholds for the relative vorticity and the warm core intensity based on the historical tropical cyclone data includes: Based on the relative vorticity and the warm core intensity, the historical relative vorticity value and historical warm core structure value within the 2° spherical neighborhood of the historical tropical cyclone center are extracted from the historical tropical cyclone data using statistical analysis. Calculate the historical relative vorticity maximum and historical warm core structure maximum for each historical tropical cyclone based on the historical relative vorticity value and the historical warm core structure value. The relative thresholds are obtained by calculating the 95th percentiles of the historical relative vorticity maxima and the historical warm-core structure maxima, respectively. The relative threshold satisfies the following formula: This is a relative threshold. For the candidate values ​​of the variable, For indicator functions, For the first The historical relative vorticity maximum or historical warm core structure maximum of a historical tropical cyclone. The number of all historical tropical cyclones used to construct the empirical distribution; Meteorological data at different times are obtained based on the numerical weather prediction model, and candidate areas at different times are obtained based on the relative threshold and the meteorological data. Based on the candidate regions at different times, multiple sets of latitude, longitude, and intensity data of tropical cyclone centers at different times were obtained; Based on the latitude and longitude data and the intensity data, a clustering algorithm is used to obtain the path and intensity information of each tropical cyclone over time.

2. The automatic tropical cyclone tracking method based on relative threshold according to claim 1, characterized in that, The relative vorticity and warm core intensity of historical tropical cyclones obtained from the historical meteorological data include: Extract the meridional and zonal wind components at a distance of 10m above the ground from the historical meteorological data; The relative vorticity of the tropical cyclone is calculated based on the meridional wind component and the zonal wind component. The first and second average temperatures of the 200 hPa isobaric surface and the 500 hPa isobaric surface were extracted from the historical meteorological data, respectively. The warm core intensity of the tropical cyclone is calculated based on the first average temperature and the second average temperature.

3. The automatic tropical cyclone tracking method based on relative thresholds according to claim 1, characterized in that, The step of obtaining candidate regions at different times based on the relative threshold and the meteorological data includes: The relative threshold is used to set constraints, and the meteorological data is filtered using the constraints to obtain candidate areas at different times. The candidate regions satisfy the following formula: in, This represents the relative threshold for relative vorticity. Relative vorticity in meteorological data. The relative threshold for the intensity of warmth. The intensity of warmth in meteorological data.

4. The automatic tropical cyclone tracking method based on relative threshold according to claim 1, characterized in that, The intensity data includes the maximum 10m wind speed and the minimum sea level pressure. The multiple sets of latitude and longitude data and intensity data of tropical cyclone centers obtained from the candidate regions at different times include: Based on the candidate areas at different times, the 10m wind speed at different times is calculated using the meteorological data. Within the candidate region, determine the vorticity maximum point, and using the vorticity maximum point as the center, determine the wind speed minimum points at different times within a 2° spherical neighborhood using the 10m wind speed. Based on the minimum wind speed points at different times, the latitude and longitude data of the tropical cyclone center at different times were obtained; The maximum 10m wind speed and minimum sea level pressure were calculated using the meteorological data within a 2° spherical neighborhood of the center of the tropical cyclone at different times.

5. The automatic tropical cyclone tracking method based on relative thresholds according to claim 1, characterized in that, The step of obtaining the time-varying path and intensity information of each tropical cyclone using a clustering algorithm based on the latitude and longitude data and the intensity data includes: A three-dimensional feature vector is constructed for each candidate region based on the latitude and longitude data and the intensity data; Construct the adjacency structure of the candidate region based on the three-dimensional feature vector; The center nodes of the candidate regions are classified according to the adjacency structure. Based on the classification results, density connectivity clusters are constructed, and the path and intensity information of each tropical cyclone over time are obtained based on the density connectivity clusters.

6. The automatic tropical cyclone tracking method based on relative threshold according to claim 5, characterized in that, The step of constructing the adjacency structure of the candidate region based on the three-dimensional feature vector includes: The spatial distance and intensity difference between the centers of two tropical cyclones at different times are calculated based on the three-dimensional feature vectors. Based on the spatial distance and the intensity difference, a comprehensive distance metric function between the centers of two tropical cyclones at different times is constructed; The adjacency structure of the candidate regions is constructed based on the comprehensive distance metric function.

7. The automatic tropical cyclone tracking method based on relative threshold according to claim 6, characterized in that, The step of constructing the adjacency structure of the candidate region based on the comprehensive distance metric function includes: Historical track data of tropical cyclones are obtained, and historical spatial data and historical intensity data of the centers of two different historical tropical cyclones are obtained based on the historical track data. Calculate the 95th percentile of the historical spatial data and the historical intensity data respectively to obtain the historical spatial threshold and the historical intensity threshold; The cluster radius is calculated based on the historical spatial threshold and the historical intensity threshold; The adjacency structure of the candidate regions is constructed based on the clustering radius and the comprehensive distance metric function.

8. The automatic tropical cyclone tracking method based on relative threshold according to claim 7, characterized in that, The adjacency structure satisfies the following formula: in, To be based on three-dimensional feature vectors The constructed adjacency structure, To be related to three-dimensional feature vectors The nearest neighbor and the combined distance metric function is less than the cluster radius value. Other three-dimensional feature vectors, For calculation using the spherical cosine formula and Spatial distance between them For intensity weight, for and The intensity difference value, This represents the cluster radius value.

Images