A tropical cyclone track screening method, system, device, medium and product

Abstract

The application discloses a tropical cyclone path screening method, system, device, medium and product, relates to the field of ocean and weather forecast, and comprises the following steps: selecting a tropical cyclone sample, determining a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method, determining a second path similar typhoon list of the tropical cyclone sample based on a contrast learning tropical cyclone similarity evaluation model, training the contrast learning tropical cyclone similarity evaluation model by using a training data set, wherein the training data set comprises CMA tropical cyclone optimal path data set and ERA5 reanalysis data, determining a tropical cyclone list similar to the path of the tropical cyclone sample according to the first path similar typhoon list and the second path similar typhoon list, and screening an optimal tropical cyclone path from the tropical cyclone list. The application comprehensively considers various factors to evaluate the similarity degree of the tropical cyclone, and avoids misjudgment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of marine and weather forecasting, in particular to a tropical cyclone path screening method, system, device, medium and product. BACKGROUND

[0002] The study of tropical cyclone similarity evolution, that is, how tropical cyclones evolve and exhibit similar behavior patterns, has always been a major focus of meteorology. Similarity is manifested in the track, intensity, structure, and other aspects of tropical cyclones. When studying the similarity of tropical cyclones, several key aspects are usually considered.

[0003] 1. Climatic characteristics.

[0004] Geographical distribution: Tropical cyclones follow certain paths, usually influenced by sea surface temperature, Coriolis effect, and prevailing wind patterns in the region.

[0005] Seasonal trends: Most tropical cyclones occur in specific seasons in the Pacific and Atlantic basins (e.g., June to November in the western Pacific). The causes, development, and tracks of storms during this period often exhibit similar patterns. Formation conditions: The formation of tropical cyclones requires specific conditions, such as warm sea water, high humidity, and low vertical wind shear. Similar conditions lead to the formation of storms with similar characteristics.

[0006] 2. Track similarity and intensity similarity.

[0007] Analyzing the historical paths of typhoons to identify recurring tracks or areas where storms are prone to form and intensify. It is well known that the waters east of the Philippines are an important source of tropical cyclones. Intensity similarity, especially manifested in the case of tropical cyclones that rapidly intensify near the coast, usually with a central pressure drop of 30 hPa (hundred pascal) within 24 hours, leading to rapid strengthening of tropical intensity, usually such tropical cyclones will cause major disaster along the coast.

[0008] 3. Impact similarity.

[0009] By identifying similar patterns between past typhoons, better warning systems can be developed to predict storm tracks, intensity, and areas of impact. Tropical cyclones with similar development or intensity in terms of rainfall, wind damage, and storm surges will have similar impacts.

[0010] From the macro perspective of tropical cyclone similarity, the track of a tropical cyclone is influenced by a combination of atmospheric and oceanic factors. The following are the main factors recognized.

[0011] 1. Sea surface temperature.

[0012] Tropical cyclones derive their energy from warm ocean waters. Higher sea surface temperatures (typically above 26°C) provide the necessary heat and moisture for tropical cyclones to become stronger and more persistent. The presence of warm water also influences the movement and duration of cyclones. While sea surface temperature does not directly affect the direction of a cyclone, it indirectly influences its intensity and development, thereby affecting the cyclone's track.

[0013] 2. Wind shear.

[0014] Wind shear refers to the variation in wind speed and direction with height. Strong wind shear (a significant difference in wind speed and direction between lower and upper levels of the atmosphere) can weaken or disrupt the structure of a tropical cyclone. Low wind shear is conducive to the formation and intensification of cyclones. In contrast, high wind shear can weaken a cyclone, potentially changing its path or causing it to dissipate. The direction and strength of wind shear can influence the movement of a cyclone.

[0015] 3. Steering winds (prevailing winds).

[0016] Steering winds are the winds in the upper levels of the atmosphere (typically 500mb-700mb) that guide a tropical cyclone along its path. These winds vary with time, geographical location, and season. The most important steering winds for tropical cyclones are the easterlies in the tropics, which usually push the cyclones westward. However, changes in steering winds, such as those caused by the subtropical jet stream or large high-pressure systems, can cause cyclones to turn northward or eastward.

[0017] 4. Pressure systems.

[0018] Tropical cyclones typically move along the edges of high and low pressure systems. The location and strength of these pressure systems can have a significant impact on the track of a cyclone. High pressure systems act as barriers that force cyclones to move around them. They can either direct a cyclone in a certain direction or completely block its forward progress. Low pressure systems can attract tropical cyclones, causing them to change direction, speed up, or alter their path. The interaction between high and low pressure systems is crucial in determining the movement of a cyclone.

[0019] 5. Topography.

[0020] When a tropical cyclone encounters land, the interaction between the storm and the land surface can significantly alter its path and intensity. Due to the lack of warm ocean water, friction with the land surface, and reduced moisture supply, a cyclone typically weakens upon landfall. If a cyclone approaches large mountain ranges (such as the Himalayas or the Andes), it can experience deflection, weakening, or even dissipation. Mountains can either block the movement of a cyclone or cause it to turn around the sides of the mountains. Coastal features such as bays or peninsulas can alter the cyclone's path by concentrating wind patterns and storm surges, often enhancing the cyclone's impact on certain areas.

[0021] 6. Mid-latitude troughs and jets.

[0022] Once a tropical cyclone begins to move away from the tropics, the location of mid-latitude troughs and jets can influence the movement of the tropical cyclone. When a tropical cyclone enters the mid-latitudes, it can interact with low troughs and jet streams, which often cause the cyclone to turn towards higher latitudes.

[0023] 7. Atmospheric instability and convection.

[0024] Cyclones are driven by convection, which is the upward movement of warm, moist air. Atmospheric instability can promote the development of convection, providing power to the cyclone. The interaction of the cyclone with unstable atmospheric conditions can affect the movement and development of the cyclone. Areas of air rising can cause changes in wind patterns, altering the trajectory of the cyclone.

[0025] 8. Structure and size of the cyclone.

[0026] The structure of the cyclone, including its size (i.e. storm radius) and wind field distribution, can affect the cyclone's response to steering winds and pressure systems. Larger cyclones are more affected by large-scale steering winds and may take longer to respond to changes in environmental conditions, while smaller cyclones may be more unstable and can change direction more quickly.

[0027] Tropical cyclones with similar paths can reflect the similarity of their environmental conditions, evolution processes, and the equivalent of the combined effects of many influencing factors. They are widely used in tropical cyclone prediction, typhoon storm surge prediction, and other businesses. For example, in typhoon storm surge prediction, we look for historical tropical cyclones with similar paths to the sample tropical cyclone, review the development of historical tropical cyclones and the actual coastal tide and storm surge, and have a very important reference value for predicting the storm surge that may be caused by the current sample tropical cyclone.

[0028] Through the analysis of existing patents and related literature, the existing tropical cyclone similarity can be concentrated in single indicators such as path similarity and intensity similarity. In path similarity, the basic starting point is to measure the degree of similarity in the graph by giving quantitative indicators, but the above method has problems. It ignores the fact that tropical cyclones are weather-scale systems, and their trajectory development and changes are influenced by a variety of ocean and meteorological factors. The graph (trajectory) is only the final manifestation, while the above complex influencing factors are the root cause. Tropical cyclones with similar paths may have very different intensities, movement speeds, and circulation systems. If we only rely on the degree of path similarity to determine the tropical cyclone path, we may make mistakes. Therefore, we need to consider multiple factors in the similarity of tropical cyclones. SUMMARY

[0029] The application aims to provide a tropical cyclone path screening method, system, device, medium and product to solve the problem that the determination of tropical cyclone path based on path similarity is prone to misjudgment.

[0030] To achieve the above-mentioned purpose, the application provides the following solutions.

[0031] In a first aspect, the application provides a tropical cyclone path screening method, comprising the following steps.

[0032] Select a tropical cyclone sample, and determine a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method.

[0033] Determine a second path similar typhoon list of the tropical cyclone sample based on a contrastive learning tropical cyclone similarity evaluation model; the contrastive learning tropical cyclone similarity evaluation model is obtained by training a training data set; the training data set comprises a CMA tropical cyclone best path data set and ERA5 reanalysis data; the CMA tropical cyclone best path data set comprises time, intensity, latitude, longitude and central minimum pressure of a tropical cyclone; the ERA5 reanalysis data comprises surface elements and meteorological elements at different pressure layers; the surface elements comprise sea level pressure and sea surface wind; the meteorological elements comprise geopotential height, wind, air temperature, vorticity and divergence.

[0034] Determine a tropical cyclone list similar to the path of the tropical cyclone sample according to the first path similar typhoon list and the second path similar typhoon list, and screen a best tropical cyclone path from the tropical cyclone list.

[0035] In a second aspect, the application provides a tropical cyclone path screening system, comprising the following modules.

[0036] A first path similar typhoon list determination module is configured to select a tropical cyclone sample, and determine a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method.

[0037] A second path similar typhoon list determination module is configured to determine a second path similar typhoon list of the tropical cyclone sample based on a contrastive learning tropical cyclone similarity evaluation model; the contrastive learning tropical cyclone similarity evaluation model is obtained by training a training data set; the training data set comprises a CMA tropical cyclone best path data set and ERA5 reanalysis data; the CMA tropical cyclone best path data set comprises time, intensity, latitude, longitude and central minimum pressure of a tropical cyclone; the ERA5 reanalysis data comprises surface elements and meteorological elements at different pressure layers; the surface elements comprise sea level pressure and sea surface wind; the meteorological elements comprise geopotential height, wind, air temperature, vorticity and divergence.

[0038] The optimal tropical cyclone path screening module is configured to determine a list of tropical cyclones similar to the sample path of the tropical cyclone according to the first path-similar typhoon list and the second path-similar typhoon list, and screen an optimal tropical cyclone path from the list of tropical cyclones.

[0039] In a third aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the tropical cyclone path screening method described above.

[0040] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executable by a processor to implement the tropical cyclone path screening method described above.

[0041] In a fifth aspect, the present application provides a computer program product comprising a computer program, wherein the computer program is executable by a processor to implement the tropical cyclone path screening method described above.

[0042] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0043] The present application first determines a first path-similar typhoon list based on a tropical cyclone path similarity method, and then determines a second path-similar typhoon list of a sample of a tropical cyclone based on a comparative learning tropical cyclone similarity evaluation model trained based on a training data set constructed from CMA tropical cyclone optimal path data set and ERA5 reanalysis data, wherein the training data set includes different influencing factors of a tropical cyclone to represent the interaction relationship between the ocean and the atmosphere, so that the list of tropical cyclones similar to the sample path of the tropical cyclone determined according to the first path-similar typhoon list and the second path-similar typhoon list is more accurate, thereby enabling the optimal tropical cyclone path to be screened out. The present application not only considers the trajectory pattern, but also considers the interaction relationship between the ocean and the atmosphere, and comprehensively considers various factors to evaluate the similarity of the tropical cyclone, thereby avoiding misjudgment. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0045] Figure 1 A flowchart of a tropical cyclone path screening method provided by an embodiment of the present application.

[0046] Figure 2 A low-pressure and strong wind area diagram provided by an embodiment of the present application.

[0047] Figure 3 A tropical cyclone number-time and result E corresponding relationship diagram provided by an embodiment of the present application.

[0048] Figure 4 A structure diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0050] In order to make the objects, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0051] The embodiments of the present application provide a tropical cyclone path screening method, including the following steps.

[0052] S1: selecting a tropical cyclone sample, and determining a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method.

[0053] S2: determining a second path similar typhoon list of the tropical cyclone sample based on a contrast learning tropical cyclone similarity evaluation model; the contrast learning tropical cyclone similarity evaluation model is obtained by training a training data set; the training data set includes a CMA tropical cyclone best path data set and ERA5 reanalysis data; the CMA tropical cyclone best path data set includes time, intensity, latitude, longitude and central minimum pressure of the tropical cyclone; the ERA5 reanalysis data includes surface elements and meteorological elements at different pressure layers; the surface elements include sea level pressure and sea surface wind; the meteorological elements include geopotential height, wind, air temperature, vorticity and divergence.

[0054] S3: determining a tropical cyclone list similar to the path of the tropical cyclone sample according to the first path similar typhoon list and the second path similar typhoon list, and screening a best tropical cyclone path from the tropical cyclone list.

[0055] In practical applications, the application selects a tropical cyclone as a sample; determines a first path similar typhoon list of the selected sample tropical cyclone based on the calculation result of the traditional path similarity (such as buffer zone analysis, etc.), that is, the tropical cyclone path similarity method; determines a second path similar typhoon list of the selected sample tropical cyclone based on the calculation result of the contrast learning tropical cyclone similarity evaluation model; and takes the intersection of the first path similar typhoon list and the second path similar typhoon list as the final tropical cyclone list.

[0056] The tropical cyclone path similarity method can be the tropical cyclone path similarity method based on the Fréchet distance, the tropical cyclone buffer zone spatial analysis method, or the tropical cyclone path similarity degree determination method and the method described in the storage medium.

[0057] The application also designs a method for selecting, organizing and processing from a training data set, generating a training data set, and based on a contrast learning model, carrying out model training and verification, and after the training result is verified and evaluated, applying the model to actual business, and introducing the synoptic features of the subtropical high and low pressure system.

[0058] In an exemplary embodiment, S2 further comprises: processing the training data set to determine the processed training data set.

[0059] In an exemplary embodiment, the data set processing background is closely related to the application scene of tropical meteorology similarity, and the tropical cyclone similarity analysis focuses on the similarity of a period of time (such as 72 hours before landing) before and after the landing of the tropical cyclone, rather than the entire path from the generation source, which can be understood as local similarity. Based on this, the training data set is processed to determine the processed training data set, which specifically includes the following steps.

[0060] S11: Based on the tropical cyclone serial number, the tropical cyclones in the CMA tropical cyclone best path data set are processed one by one to form a path point data set; the processing includes deleting records with a sub-center identifier.

[0061] S12: Based on the tropical cyclone serial number, 500hpa geopotential height grid point data, sea level pressure data, 10m wind U component data and 10m wind V component data are extracted from the ERA5 reanalysis data according to time; wherein U is the horizontal wind speed component; V is the vertical wind speed component.

[0062] S13: Determine the subtropical high pressure contour surface according to the 500hpa geopotential height grid point data.

[0063] S14: determining a tropical cyclone low pressure area according to the sea level pressure data, the 10-meter wind U component data, and the 10-meter wind V component data.

[0064] S15: fusing the path point data set, the subtropical high pressure isosurface, and the tropical cyclone low pressure area to construct a processed training data set.

[0065] In actual application, the China Meteorological Administration (CMA) tropical cyclone best path data set is from the data of the China Meteorological Administration Tropical Cyclone Data Center, and the CMA tropical cyclone best path data set includes the position and intensity of tropical cyclones in the northwest Pacific (north of the equator and west of 180° east longitude) every 6 hours since 1949, including time, intensity, latitude, longitude, and central minimum pressure, etc., wherein the time is Coordinated Universal Time (UTC).

[0066] The ERA5 reanalysis data is from the European Centre for Medium-Range Weather Forecasts, including ERA5 hourly data on pressure levels from 1940 to present and ERA5 hourly data on single levels from 1940 to present, and specifically including meteorological elements such as geopotential height, wind, air temperature, vorticity, and divergence since 1940 at different pressure layers, and surface elements including sea level pressure and sea surface wind (10 meters) and the like.

[0067] In an exemplary embodiment, S13 can include the following steps.

[0068] S21: contour line tracking is performed on the 500hpa geopotential height grid point data with the condition of an isopleth greater than 5880gpm, and a first isopleth result is determined by using the Marching Squares algorithm.

[0069] S22: the first isopleth result is preliminarily screened to determine a preliminarily screened isopleth result.

[0070] S23: the preliminarily screened isopleth result is simplified by using the Douglas-Peucker algorithm to determine a simplified isopleth result.

[0071] S24: each point in the simplified isopleth result is sequentially connected to generate subtropical high surface data; the subtropical high surface data is a subtropical high pressure isosurface.

[0072] In an exemplary embodiment, S14 can include the following steps.

[0073] S31: Contour tracking is performed on the sea level pressure data in the interval of 910hpa to 1000hpa with an interval of 5hpa, and a second contour result is determined by using the Marching Squares algorithm.

[0074] S32: A low pressure area of the 990hpa isobar is determined according to the second contour result.

[0075] S33: Gridded wind speed data is generated according to the 10-meter wind U component data and the 10-meter wind V component data.

[0076] S34: An isobaric area of 6-level wind or above is extracted based on the gridded wind speed data; the isobaric area of 6-level wind or above is a gale area of 6-level wind or above.

[0077] S35: A low pressure gale area is determined according to the intersection area of the low pressure area and the gale area; the low pressure gale area is a tropical cyclone low pressure area.

[0078] In an exemplary embodiment, S32 can include the following steps.

[0079] S41: The second contour result is processed according to a threshold condition to determine a processed contour result.

[0080] S42: The processed contour result is simplified by using the Douglas-Peucker algorithm to determine a simplified contour result, and a low pressure area of the 990hpa isobar is selected according to the simplified contour result.

[0081] Taking the northwest Pacific region as an example, the generation and processing process of the data set are further trained.

[0082] First of all, it needs to be determined that: 1. Tropical cyclones are affected by the subtropical high pressure, and the spatial distribution of the subtropical high pressure has a significant impact on the movement of tropical cyclones.

[0083] 2. Low pressure systems will attract tropical cyclones, causing them to change direction, speed up or change path.

[0084] 3. According to 1 and 2, it can be known that the interaction between high and low pressure systems is crucial to determining the movement of tropical cyclones.

[0085] In addition, based on the idea of local similarity, the tropical cyclones will be truncated in the data set processing, which is used to expand the number of data sets on the one hand, and to focus on the local similarity on the other hand.

[0086] Extract tropical cyclones that affect the sea area of China or land within 72 hours (the definition of affecting the sea area of China is entering the 72-hour warning line).

[0087] The same tropical cyclone is recalculated in 6-hour intervals to generate a training data set.

[0088] The training data set generation and processing process is as follows.

[0089] ①CMA tropical cyclone best path data set, processed one by one according to tropical cyclone serial number, delete sub-center case (processing method is to delete the record with sub-center mark), form path point data set including tropical cyclone serial number, time, longitude, latitude, intensity, central minimum pressure, the time interval between path points is 6h (hours).

[0090] ②According to its time (UTC), extract from ERA5 hourly data on pressure levels from 1940 to present and ERA5 hourly data on single levels from 1940 to present two data sets, respectively extract 500hpa geopotential height and sea level pressure, 10m wind (U, V) four data sets (grid points); Considering the spatial range matching of CMA tropical cyclone path set (Northwest Pacific region), the spatial range of the above two ERA5 data sets is limited to a certain latitude and longitude range, the spatial range selected in this paper is [north latitude 0-55°, east longitude 90-180°].

[0091] ③500hpa geopotential height grid point data extracts the contour line greater than 5880gpm, obtains the subtropical high contour surface, the specific steps are as follows.

[0092] a) Based on 500hpa geopotential height grid point data, 5880gpm as the condition, carry out contour line tracking, generate using Marching Squares algorithm, set its result as the first contour line result A.

[0093] b) Preliminary screening for A, delete small area closed curve (contour line less than 6 points), isolated point and other cases, after the above processing, get the preliminary screening contour line result B.

[0094] c) For B, the Douglas-Peucker algorithm is used to simplify the generated contour, and the preset threshold value is 3-4.5 times the shortest distance in B (i.e. the Euclidean distance between two adjacent points in B). After Douglas-Peucker algorithm processing, the simplified contour result C is obtained, that is, the 5880gpm value contour data.

[0095] d) For C, the points in the contour are connected in turn to generate the corresponding isosurface, that is, the subtropical high pressure isosurface data.

[0096] (4) The sea level pressure, 10-meter wind U component, and 10-meter wind V component grid data are extracted from the tropical cyclone low pressure area, and at the same time, the low pressure area is verified through the gale area (6 or more). The specific steps are as follows.

[0097] a) Based on the sea level pressure grid data, an isobar is generated with a preset interval of 5hpa from 910hpa to 1000hpa. Marching Squares algorithm is used for isobar tracking and generation. Set the result as the second isobar result A1, which includes multiple isobars.

[0098] b) For A1, a threshold is used to filter each isobar. Small area closed curves (isobars less than 6 points) and isolated points are deleted. After the above processing, the processed isobar result B1 is obtained, which includes multiple isobars.

[0099] c) For B1, the Douglas-Peucker algorithm is used to simplify the generated isobar, and the preset threshold value is 1.5-2.5 times the shortest distance in B1 (i.e. the Euclidean distance between two adjacent points in B1). After Douglas-Peucker algorithm processing, the simplified isobar result C is obtained.

[0100] d) Based on the 10-meter wind U component and 10-meter wind V component grid data, the grid wind speed data is calculated , the calculation formula is .

[0101] The threshold value of 6 or more gale is set to 10.8-13.8m / s, and the value of the present application is 12m / s.

[0102] According to the isobar generation method shown in ③, the isobar area of 6-level wind is extracted, and the isosurface is finally generated.

[0103] e) According to the spatial range of step c) (selecting the low pressure area of 990hpa isobar), and the coverage area of step d) (6 or more wind), the spatial intersection operation provided by GIS software is used to obtain the intersection range of the two.

[0104] f) After the above operations are completed, the face feature data of the low-pressure strong wind area is obtained, denoted as low-pressure strong wind area D, as shown in Figure 2 .

[0105] ⑤ The subtropical high pressure isosurface and the low-pressure strong wind area D obtained in ③ and ④ are fused based on space, that is, the two images are merged according to the corresponding latitude and longitude, and the merged result E.

[0106] After the data processing is performed through the above steps, the data set is organized in the following structure. The tropical cyclone number-time is used as the unique identification code, which is associated with the result E of step ⑤ corresponding to the time, and a corresponding relationship is established, as shown in Figure 3 .

[0107] In practical applications, the contrast learning tropical cyclone similarity evaluation model can be a simple twin network (SimpleTwin Network Simsiam, Simsiam) model, and the enhancement process of the Simsiam model is customized.

[0108] The present application enhances the merged result E through the following two branches.

[0109] Branch one is to enhance the merged result E through the Canny algorithm. The Canny edge detection algorithm has four steps, including the following: 1) Gaussian filtering or median filtering is performed on the image to filter noise; 2) the Sobel operator is used to calculate the gradient size and direction of each pixel point of the image; 3) the non-maximum suppression algorithm is used to select the best edge in a group of edges, which specifically checks each pixel point and the nearby pixel points with consistent gradient direction, and the current pixel point gradient is the largest, then it is retained, otherwise it is removed; 4) the double threshold is used to determine the final edge, the pixel point gradient is higher than the large threshold, then it is retained; the pixel point is lower than the small threshold, then it is ignored; between the two thresholds, it is judged whether the pixel point is connected with the edge pixel point.

[0110] After the above processing, the enhanced result X1 of the merged result E can be obtained.

[0111] Branch two is to intercept a local image in the merged result E and perform binaryzation processing. The specific processing method is to take the center of the tropical cyclone low pressure center as the center point, intercept the image along the southeast, southwest and northwest directions with a threshold of 5°, and perform binaryzation processing to generate the enhanced result X2.

[0112] The SimSiam model also includes an encoder, an MLP, a loss function, etc., which will not be described here.

[0113] In addition, data from 1949 to 2023 is selected for model training, and the training data set is divided into a training set and a test set in a ratio of 7:3.

[0114] The present application not only considers trajectory patterns, but also considers the interaction relationship of marine atmosphere, and evaluates the similarity of tropical cyclones by comprehensively considering various factors, so that misjudgment is avoided.

[0115] Based on the same inventive concept, the present application also provides a tropical cyclone path screening system. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme described in the above method, so the specific limitations in one or more tropical cyclone path screening system embodiments provided below can refer to the limitations of the tropical cyclone path screening method in the foregoing, which will not be repeated here.

[0116] In an exemplary embodiment, a tropical cyclone path screening system is provided, comprising the following modules.

[0117] A first path similar typhoon list determination module is configured to select a tropical cyclone sample and determine a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method.

[0118] A second path similar typhoon list determination module is configured to determine a second path similar typhoon list of the tropical cyclone sample based on a contrastive learning tropical cyclone similarity evaluation model; the contrastive learning tropical cyclone similarity evaluation model is obtained by training a training data set; the training data set includes a CMA tropical cyclone best path data set and ERA5 reanalysis data; the CMA tropical cyclone best path data set includes time, intensity, latitude, longitude and central minimum pressure of a tropical cyclone; the ERA5 reanalysis data includes surface elements and meteorological elements at different pressure layers; the surface elements include sea level pressure and sea surface wind; the meteorological elements include geopotential height, wind, air temperature, vorticity and divergence.

[0119] A best tropical cyclone path screening module is configured to determine a tropical cyclone list similar to the path of the tropical cyclone sample according to the first path similar typhoon list and the second path similar typhoon list, and screen a best tropical cyclone path from the tropical cyclone list.

[0120] In an exemplary embodiment, a computer device is provided, which includes a memory and a processor, the memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program. The computer device can be a server or a terminal, and its internal structure diagram can be as follows Figure 4As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through the system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capability. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data to be processed. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with the terminal outside through the network connection. The computer program is executed by the processor to realize a tropical cyclone path screening method.

[0121] Those skilled in the art can understand that, Figure 4 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement. In one exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to realize the steps in each of the above method embodiments.

[0122] In one exemplary embodiment, a computer readable storage medium is provided, storing a computer program, which is executed by a processor to realize the steps in each of the above method embodiments.

[0123] In one exemplary embodiment, a computer program product is provided, including a computer program, which is executed by a processor to realize the steps in each of the above method embodiments.

[0124] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0125] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to a memory, a database or other medium used in the embodiments provided in the present application can include at least one of a non-volatile and a volatile memory. The non-volatile memory can include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical storage, a high-density embedded non-volatile memory, a resistive random access memory (ReRAM), a magnetoresistive random access memory (MRAM), a ferroelectric random access memory (FRAM), a phase change memory (PCM), a graphene memory, etc. The volatile memory can include a random access memory (RAM) or an external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc.

[0126] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0127] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0128] The principles and implementation modes of the present application are described by using specific examples in the present application. The above embodiments are only used to help understand the method and its core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range can be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method of tropical cyclone track screening, the method comprising: include: Select a tropical cyclone sample and, based on the tropical cyclone path similarity method, determine a first path similarity list of typhoons for the tropical cyclone sample; Based on the comparative learning tropical cyclone similarity assessment model, a second path similarity list of typhoons for the tropical cyclone sample is determined. The comparative learning tropical cyclone similarity assessment model is trained using a training dataset. This training dataset includes the CMA tropical cyclone optimal path dataset and ERA5 reanalysis data. The CMA tropical cyclone optimal path dataset includes the time, intensity, latitude, longitude, and minimum central pressure of the tropical cyclone. The ERA5 reanalysis data includes surface elements and meteorological elements at different pressure levels. The surface elements include sea level pressure and sea surface wind. The meteorological elements include geopotential height, wind, temperature, vorticity, and divergence. Based on a comparative learning tropical cyclone similarity assessment model, a second list of similar typhoons with similar paths to the tropical cyclone sample was determined, which previously included: The training dataset is processed to determine the processed training dataset. Specifically, this includes: processing each tropical cyclone in the CMA tropical cyclone optimal path dataset based on the tropical cyclone number to form a path point dataset; the processing includes deleting records with sub-center identifiers. Based on the tropical cyclone sequence number, and according to time, the 500 hpa geopotential height grid data, sea level pressure data, 10-meter wind U-component data, and 10-meter wind V-component data are extracted from the ERA5 reanalysis data; where U is the horizontal wind speed component and V is the vertical wind speed component. Based on the 500 hPa geopotential height grid data, the isosurface of the subtropical high is determined, specifically including: Using contour lines greater than 5880 gpm as a condition, contour line tracing is performed on the 500 hpa geopotential height grid data, and the Marching Squares algorithm is used to determine the first contour line result. The first contour line results are initially screened to determine the contour line results after the initial screening. The Douglas-Peucker algorithm is used to simplify the contour line results after the initial screening, and the simplified contour line results are determined. Connect the points in the simplified contour lines sequentially to generate subtropical high surface data; the subtropical high surface data is the subtropical high pressure isosurface. Based on the sea level pressure data, the 10-meter wind U-component data, and the 10-meter wind V-component data, the tropical cyclone low-pressure area is determined with an interval of 910 hpa to 1000 hpa and an interval of 5 hpa. The sea level pressure data is traced by isolines, and the Marching Squares algorithm is used to determine the second isoline result. The low-pressure region of the 990 hPa isobaric line is determined based on the results of the second contour line. Based on the 10-meter wind U-component data and the 10-meter wind V-component data, gridded wind speed data is generated. Based on the gridded wind speed data, isoline areas with winds of force 6 or above are extracted; these isoline areas with winds of force 6 or above are areas with strong winds of force 6 or above. Determine a low-pressure gale area according to an intersection area of the low-pressure area and the gale area; the low-pressure gale area is a tropical cyclone low-pressure area; Fuse the path point data set, the subtropical high pressure contour surface, and the tropical cyclone low-pressure area to construct a processed training data set; Determine a tropical cyclone list similar to the tropical cyclone sample path according to the first path similar typhoon list and the second path similar typhoon list, and screen an optimal tropical cyclone path from the tropical cyclone list.

2. The tropical cyclone track screening method of claim 1, wherein, Determine a low-pressure area of a 990hpa isobaric line according to the second contour result, specifically including: Process the second contour result according to a threshold condition to determine a processed contour result; Simplify the processed contour result by using a Douglas-Peucker algorithm to determine a simplified contour result, and select a low-pressure area of a 990hpa isobaric line according to the simplified contour result.

3. A tropical cyclone track screening system characterized by, The tropical cyclone path screening system adopts the tropical cyclone path screening method of any one of claims 1-2, and includes: A first path similar typhoon list determination module configured to select a tropical cyclone sample and determine a first path similar typhoon list of the tropical cyclone sample based on a tropical cyclone path similarity method; A second path similar typhoon list determination module configured to determine a second path similar typhoon list of the tropical cyclone sample based on a contrast learning tropical cyclone similarity evaluation model; the contrast learning tropical cyclone similarity evaluation model is trained by a training data set; the training data set includes a CMA tropical cyclone optimal path data set and ERA5 reanalysis data; the CMA tropical cyclone optimal path data set includes time, intensity, latitude, longitude, and central minimum pressure of a tropical cyclone; the ERA5 reanalysis data includes surface elements and meteorological elements at different pressure layers; the surface elements include sea level pressure and sea surface wind; the meteorological elements include geopotential height, wind, air temperature, vorticity, and divergence; An optimal tropical cyclone path screening module configured to determine a tropical cyclone list similar to the tropical cyclone sample path according to the first path similar typhoon list and the second path similar typhoon list, and screen an optimal tropical cyclone path from the tropical cyclone list.

4. A computer device comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the tropical cyclone path screening method of any one of claims 1-2.

5. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the tropical cyclone path screening method of any one of claims 1-2.

6. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the tropical cyclone path screening method of any one of claims 1-2.

Images