A typhoon intensity prediction method, system, device and medium
By using ERA5 reanalysis data and Fuxi meteorological model combined with a hybrid model of 3D convolutional neural network and long and short-term memory network, the problem of error accumulation and insufficient small-scale structural forecasting in typhoon intensity prediction is solved, and high-precision typhoon intensity prediction is achieved.
Patent Information
- Application Number
- CN202510195563.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-21
- Publication Date
- 2025-07-04
- Estimated Expiration
- 2045-02-21
AI Technical Summary
Existing AI meteorological models have problems of accumulation of errors and insufficient small-scale structural forecasting in the prediction of typhoon intensity, resulting in insufficient prediction accuracy and reliability, especially in poor performance in extreme typhoon events.
ERA5 reanalysis data and Fuxi meteorological model are used, and a hybrid model of 3D convolutional neural network and long-term memory network is combined. By acquiring and correcting the channel dimension data of the central location of the typhoon, time dimension data is formed to output typhoon intensity prediction at future moments.
It improves the accuracy and reliability of typhoon intensity prediction, especially in extreme events, which significantly improves the timeliness and accuracy of predictions.
Smart Images
Figure CN119667824B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of meteorological forecasting, and specifically relates to a typhoon intensity forecasting method, system, equipment and medium. Background Art
[0002] Typhoon, as a kind of strong tropical cyclone weather phenomenon, its prediction accuracy is of vital importance for disaster prevention and mitigation. In recent years, with the rapid development of artificial intelligence (AI) technology, AI meteorological big models have been widely used in the field of typhoon prediction. However, the current AI meteorological big models generally have the problem of weak intensity prediction when predicting typhoon intensity, which is mainly due to the accumulation of errors in the model in dealing with long time scales and insufficient prediction of small-scale structures. When performing long-term typhoon prediction tasks, the existing AI models gradually accumulate errors, resulting in a gradual increase in the deviation between the prediction results and the actual observation data. This phenomenon is particularly prominent in typhoon intensity prediction, which greatly affects the accuracy and reliability of the prediction. On the other hand, the lack of prediction ability of small-scale structures is also a major problem, which is reflected in the core areas of typhoons, such as the typhoon eye and eyewall. These areas contain complex small-scale structures and play a decisive role in the changes in typhoon intensity. The current AI models have obvious deficiencies in capturing and predicting these small-scale structures, resulting in limited prediction ability of typhoon intensity, especially poor performance in extreme typhoon events.
[0003] Compared with AI models, traditional typhoon intensity prediction methods mainly rely on tropical cyclone (TC) tracking algorithms to track and predict the development of typhoons through the constraints of model meteorological variables (such as wind field, pressure field, temperature field, etc.). However, the effects of these methods on intensity prediction are not ideal. The main reasons are as follows: First, the limitations of model meteorological variables limit the accuracy of prediction. Traditional methods only rely on limited meteorological variables and fail to make full use of the correlation between multi-source data, thus affecting the comprehensiveness of prediction. Secondly, the lack of ability to capture small-scale structures is also a major shortcoming. Traditional methods have limited accuracy in simulating small-scale structures in the core area of typhoons, resulting in large deviations in the prediction of typhoon intensity. Finally, the adaptability to extreme events is poor. In extreme typhoon events, the prediction ability of traditional methods has declined significantly, making it difficult to meet the actual needs of disaster prevention and mitigation. The physical constraints of meteorological-related variables are also a key influencing aspect. Summary of the invention
[0004] Based on the above-mentioned shortcomings and deficiencies in the prior art, one of the objects of the present invention is to solve at least one or more of the above-mentioned problems in the prior art. In other words, one of the objects of the present invention is to provide a typhoon intensity prediction method, system, equipment and medium that meet one or more of the above-mentioned needs, so as to achieve the purpose of improving the accuracy and reliability of typhoon intensity prediction.
[0005] To achieve the above-mentioned invention objectives, the present invention adopts the following technical solutions:
[0006] In a first aspect, the present invention provides a typhoon intensity prediction method, including the steps of: S1, obtaining typhoon initial field data, where the typhoon initial field data includes ERA5 reanalysis data at at least two prior moments; S2, obtaining the Fuxi meteorological large model and inputting the typhoon initial field data thereto to output typhoon prediction initial data at a subsequent moment, where the typhoon prediction initial data includes the typhoon center position and typhoon characterization variables, and the typhoon characterization variables include basic field variables, thermal field variables, air pressure, and precipitation; S3, calculating other characterization variables based on the output typhoon characterization variables, where the other characterization variables include diagnostic field variables and wind speed, and collectively referring to the basic field variables, thermal field variables, air pressure, precipitation, diagnostic field variables, and wind speed as channel dimension data; S4, selecting the channel dimension data within a preset diameter range with the typhoon center position as a reference to obtain spatial dimension data; S5, obtaining preset timing parameters, arranging the spatial dimension data in chronological order to form time dimension data, and inputting the time dimension data into the obtained systematic deviation correction model to output corrected typhoon intensity prediction data at a future moment.
[0007] As a preferred solution, the basic field variables include wind field components, wind field components, and geopotential height; the thermal field variables include temperature and relative humidity; the diagnostic field variables include vorticity field, divergence field, potential vorticity, and kinetic energy.
[0008] As a preferred solution, in step S4, selecting the channel dimension data within a preset diameter range with the typhoon center position as a reference to obtain spatial dimension data specifically means: taking the typhoon center position as the center, selecting the channel dimension data with a horizontal resolution of 0.25°×0.25° and a spatial range of a diameter of 10 degrees to obtain spatial dimension data; when the spatial range is a diameter of 10 degrees, the corresponding pixels are 41×41.
[0009] As a preferred solution, the preset input timing parameter is 24 hours, the time interval is 6 hours, and there are a total of 4 time steps; the time dimension data includes each spatial dimension data corresponding to each time step.
[0010] As a preferred solution, the systematic deviation correction model adopts a hybrid model of a 3D convolutional neural network and a long short-term memory network. Step S5 is specifically as follows: input the spatial dimension data corresponding to the corresponding time into the input layer of the 3D convolutional neural network to extract the dynamic field feature, the thermodynamic field feature, and the ground field feature; obtain the path prediction data and the intensity prediction data output by the output layer of the 3D convolutional neural network; obtain a preset loss function and correct the path prediction data and the intensity prediction data respectively to obtain the typhoon intensity prediction data for the next 24 hours and / or 48 hours and / or 72 hours.
[0011] As a preferred solution, the calculation formula of the loss function is:
[0012] , where is the mean square error loss, is the mean absolute error loss, is the physical constraint loss, are the loss coefficients of the mean square error, the mean absolute error loss, and the physical constraint respectively.
[0013] As a preferred solution, calculating other characterization variables based on the output typhoon characterization variables in step S3 includes calculating the horizontal component of the divergence field, the vertical component of the vorticity field, and the kinetic energy based on the component of the wind field and the component of the wind field; the calculation formula for the horizontal component of the divergence field is , where represents the horizontal component of the divergence field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west direction and the north-south direction respectively; the calculation formula for the vertical component of the vorticity field is , where represents the vertical component of the vorticity field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west direction and the north-south direction respectively; the calculation formula for the kinetic energy is , where represents the kinetic energy, is the component of the wind field, is the component of the wind field.
[0014] In a second aspect, the present invention provides a typhoon intensity prediction system for implementing the typhoon intensity prediction method as described in the first aspect.
[0015] In a third aspect, the present invention provides an electronic device, which includes a memory, a processor, and a computer program. When the computer program is executed by the processor, it implements the typhoon intensity prediction method as described in the first aspect.
[0016] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by the processor, it implements the typhoon intensity prediction method as described in the first aspect.
[0017] Compared with the prior art, the present invention has the following beneficial effects:
[0018] 1. Using ERA5 reanalysis data as the typhoon initial field data, it has high spatio-temporal resolution and rich meteorological element information, providing a reliable data basis for the prediction of typhoon intensity.
[0019] 2. Utilizing the powerful data processing and pattern recognition capabilities of the Fuxi meteorological large model, it can accurately capture the development and evolution laws of typhoons, improving the accuracy and timeliness of predictions.
[0020] 3. First, use the Fuxi meteorological large model to output some variables characterizing typhoon intensity, and then further calculate other variables based on the already output ones to achieve the comprehensive construction of channel dimension data, ensuring the comprehensiveness and accuracy of the data, and providing rich information support for subsequent predictions.
[0021] 4. Taking the typhoon center position as a reference, select the channel dimension data within a preset diameter range to form spatial dimension data. This step ensures the pertinence and accuracy of the prediction results and avoids the interference of irrelevant information.
[0022] 5. Obtain the preset timing parameters, arrange the spatial dimension data in chronological order to form time dimension data, and input it into the systematic deviation correction model for prediction. The reasonable organization of time dimension data helps to capture the change trend of typhoon intensity and improve the accuracy of prediction.
[0023] In summary, the typhoon intensity prediction method of the present invention realizes the accurate prediction of typhoon intensity through the application of high-precision data sources, the efficient utilization of the Fuxi meteorological large model, the comprehensive construction of channel dimension data, the accurate extraction of spatial dimension data, the reasonable organization of time dimension data, and the application of the systematic deviation correction model.
[0024] Furthermore, more detailed beneficial effects will be described in combination with specific embodiments in the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0026] Figure 1 It is a schematic flowchart of the typhoon intensity prediction method provided by the embodiments of the present invention.
[0027] Figure 2 It is a structural diagram of the electronic device provided by the embodiments of the present invention.
[0028] Figure 3 It is a result graph of the comparative test provided by the embodiments of the present invention.
[0029] Figure 4 It is a result graph of the comparative test provided by the embodiments of the present invention.
[0030] Figure 5 It is a result graph of the comparative test provided by the embodiments of the present invention.
[0031] Figure 6 It is a result graph of the comparative test provided by the embodiments of the present invention.
[0032] Reference numerals in the drawings:
[0033] 200, electronic device;
[0034] 201, processor; 202, communication bus; 203, user interface; 204, network interface; 205, memory. Detailed implementation manners
[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention.
[0036] In the following introduction, multiple embodiments of the present invention are provided. Different embodiments can be replaced or combined. Therefore, the present invention can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing all other possible combinations of A, B, C, and D, even though such embodiments may not be explicitly described in the following text.
[0037] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the elements described without departing from the scope of the present disclosure. Various processes or components may be appropriately omitted, substituted, or added to each example. For example, the methods described may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with respect to some examples may be combined into other examples.
[0038] To facilitate a better understanding of the embodiments of the present invention, before explaining the specific embodiments of the present invention in detail, its application scenarios will be described first.
[0039] The typhoon intensity prediction method described in the embodiments of this specification is applied to the process of typhoon monitoring and early warning. In these scenarios, the application of the typhoon intensity prediction method aims to enable the meteorological department to more accurately predict the intensity change trend of typhoons, so that relevant departments and the public can take effective defense measures in advance to provide timely and reliable typhoon early warning information and reduce the losses caused by typhoon disasters to life and property.
[0040] Embodiment 1:
[0041] As Figure 1 shown, this embodiment provides a typhoon intensity prediction method, including the steps of: S1, obtaining typhoon initial field data, where the typhoon initial field data includes ERA5 reanalysis data at at least two prior times; S2, obtaining the Fuxi meteorological large model and inputting the typhoon initial field data thereto to output typhoon prediction initial data at a subsequent time, where the typhoon prediction initial data includes the typhoon center position and typhoon characterization variables, and the typhoon characterization variables include basic field variables, thermal field variables, air pressure, and precipitation; S3, calculating other characterization variables based on the output typhoon characterization variables, where the other characterization variables include diagnostic field variables and wind speed, and the basic field variables, thermal field variables, air pressure, precipitation, diagnostic field variables, and wind speed are collectively referred to as channel dimension data; S4, selecting the channel dimension data within a preset diameter range based on the typhoon center position to obtain spatial dimension data; S5, obtaining preset time series parameters, arranging the spatial dimension data in chronological order to form time dimension data, and inputting the time dimension data into the obtained systematic deviation correction model to output corrected typhoon intensity prediction data at a future time.
[0042] Specifically, this embodiment provides a preferred implementation manner, where the basic field variables include wind field components, wind field components, and geopotential height; the thermal field variables include temperature and relative humidity; the diagnostic field variables include vorticity field, divergence field, potential vorticity, and kinetic energy.
[0043] More specifically, the wind field The total number of components is 14, and their respective variable names are 10, 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000. More specifically, the wind field The total number of components is 14, and their respective variable names are 10, 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000. The total number of geopotential heights is 13, and their respective variable names are 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000. The total number of temperatures is 14, and their respective variable names are 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000, 2 m. The total number of the relative humidity is 13, and its respective variable names are 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000. More specifically, the variable name of the precipitation is . The total number of the vorticity field is 3, and its respective variable names are 200, 500, 850. The total number of the divergence field is 2, and its respective variable names are 200 and 850. The total number of the potential vorticity is 2, and its respective variable names are 200 and 850. The total number of the kinetic energy is 2, and its respective variable names are 300 and 500. More specifically, the variable name of the air pressure is . More specifically, the variable name of the wind speed is 10.
[0044] It can be understood that the wind field ( component and component) can describe the atmospheric motion state, indicate the typhoon circulation intensity, and reflect the typhoon movement direction. Among them, the 200 hPa divergence and the 850 hPa convergence are particularly important for typhoon path prediction. The geopotential height ( ) can characterize the atmospheric pressure distribution and reflect the typhoon system intensity. The temperature ( ) can describe the atmospheric thermal state, and the sea surface temperature is particularly crucial for typhoon generation and development. The relative humidity can characterize the water vapor content, determine the precipitation intensity, and affect the typhoon maintenance and development. The vorticity field ( ) can describe the cyclonic rotation. The positive vorticity center corresponds to the typhoon center, the vorticity intensity reflects the typhoon intensity, and the vorticity advection affects the typhoon movement. The divergence field ( ) can characterize the air flow convergence and divergence and affect the typhoon intensity change. The potential vorticity ( Comprehensively reflecting the thermal characteristics, the potential vorticity anomaly is related to the typhoon intensity, and the potential vorticity conservation affects the typhoon movement. Kinetic energy ( ) can describe the intensity of motion, reflect the intensity of the typhoon circulation, indicate the strength of the wind field, and embody the energy of the system. In summary, each of the above variables is of great significance for typhoon intensity prediction. Therefore, in this embodiment, 10-channel data are calculated from the 70-channel data output by the Fuxi meteorological large model, ensuring the comprehensiveness and accuracy of the data, and can significantly improve the accuracy and timeliness of typhoon intensity and path prediction.
[0045] Specifically, this embodiment provides a preferred implementation manner. The step S4 of selecting the channel dimension data within a preset diameter range based on the typhoon center position to obtain the spatial dimension data is specifically: centered on the typhoon center position, selecting the channel dimension data with a horizontal resolution of 0.25°×0.25° and a spatial range of a diameter of 10 degrees to obtain the spatial dimension data; the corresponding pixels when the spatial range is a diameter of 10 degrees are 41×41.
[0046] Specifically, this embodiment provides a preferred implementation manner. The preset input time series parameter is 24 hours, the time interval is 6 hours, and there are a total of 4 time instances; the time dimension data includes each spatial dimension data corresponding to each time instance.
[0047] Specifically, this embodiment provides a preferred implementation manner. The systematic deviation correction model adopts a hybrid model of a 3D convolutional neural network and a long short-term memory network. Step S5 is specifically as follows: Input the spatial dimension data corresponding to the corresponding time into the input layer of the 3D convolutional neural network to extract the dynamic field feature, thermal field feature, and ground field feature; Obtain the path prediction data and intensity prediction data output by the output layer of the 3D convolutional neural network; Obtain a preset loss function and correct the path prediction data and the intensity prediction data respectively to obtain the typhoon intensity prediction data for the next 24 hours and / or 48 hours and / or 72 hours. It can be understood that, in order to further improve the prediction accuracy, this embodiment uses a long short-term memory network (LSTM) to process these prediction data. The LSTM model can capture the long-term and short-term dependencies in the data and has special advantages for time series data (such as the changes in typhoon path and intensity over time). Therefore, after obtaining the output of the 3D convolutional neural network, these prediction data will be further input into the LSTM model to capture the dynamic change features in the time series. Next, obtain a preset loss function and correct the path prediction data and intensity prediction data respectively. The loss function is used to measure the difference between the model prediction result and the true result, and the model parameters can be continuously optimized by minimizing the loss function. In this embodiment, the loss function is used to correct the deviations in the path prediction data and intensity prediction data, so as to obtain a more accurate prediction result. Finally, the corrected prediction data can reflect the typhoon intensity changes in the next 24 hours, 48 hours, or 72 hours. This hybrid model combines the advantages of the 3D convolutional neural network in feature extraction and the ability of the LSTM model in processing time series data, and can significantly improve the accuracy and reliability of typhoon intensity prediction. Therefore, adopting a hybrid model of a 3D convolutional neural network and a long short-term memory network (LSTM) can make full use of the advantages of the two models and improve the accuracy and timeliness of typhoon intensity prediction.
[0048] Specifically, this embodiment provides a preferred implementation manner. The calculation formula of the loss function is as follows:
[0049] , where is the mean squared error loss, is the mean absolute error loss, is the physical constraint loss, are the loss coefficients of the mean squared error, mean absolute error loss, and physical constraint respectively.
[0050] Specifically, this embodiment provides a preferred implementation manner. Calculating other characterization variables based on the output typhoon characterization variables in step S3 includes based on the wind field component and the wind field Calculate the horizontal component of the divergence field, the vertical component of the vorticity field, and the kinetic energy; the formula for calculating the horizontal component of the divergence field is , where represents the horizontal component of the divergence field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west and north-south directions respectively; the formula for calculating the vertical component of the vorticity field is , where represents the vertical component of the vorticity field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west and north-south directions respectively; the formula for calculating the kinetic energy is , where represents the kinetic energy, is the component of the wind field, is the component of the wind field.
[0051] Example 2:
[0052] This example provides a typhoon intensity prediction system for implementing the typhoon intensity prediction method as described in Example 1.
[0053] Example 3:
[0054] As shown in Figure 2 , this example provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.
[0055] Among them, the communication bus can be used to realize the connection and communication of the above-mentioned various components.
[0056] Among them, the user interface may include keys, and the optional user interface may further include a standard wired interface and a wireless interface.
[0057] Among them, the network interface may but is not limited to including a Bluetooth module, an NFC module, a Wi-Fi module, etc.
[0058] Among them, the processor may include one or more processing cores. The processor uses various interfaces and circuits to connect various parts within the entire electronic device, and by running or executing instructions, programs, code sets, or instruction sets stored in the memory, as well as calling data stored in the memory, it executes various functions of the electronic device and processes data. Optionally, the processor may be implemented in at least one hardware form of DSP, FPGA, or PLA. The processor may integrate one or a combination of several of CPU, GPU, and modem, etc. Among them, the CPU mainly processes the operating system, user interface, application programs, etc.; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communication. It can be understood that the above-mentioned modem may not be integrated into the processor and can be implemented separately through a single chip.
[0059] Among them, the memory may include RAM and may also include ROM. Optionally, the memory includes a non-transitory computer-readable medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area. Among them, the program storage area can store instructions for implementing the operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area can store the data involved in the above-mentioned various method embodiments. Optionally, the memory may also be at least one storage device located far from the aforementioned processor. As a computer storage medium, the memory may include an operating system, a network communication module, a user interface module, and a typhoon intensity prediction application program. The processor can be used to call the typhoon intensity prediction application program stored in the memory and execute the steps of typhoon intensity prediction mentioned in the foregoing embodiments.
[0060] Embodiment 4:
[0061] This embodiment provides a computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When they run on a computer or a processor, they cause the computer or the processor to execute one or more steps in the above-mentioned embodiments. If the various component modules of the above-mentioned electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.
[0062] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this specification are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a Digital Versatile Disc (DVD)), or a semiconductor medium (for example, a Solid State Disk (SSD)), etc.
[0063] Embodiment Five:
[0064] To verify the effectiveness of the typhoon intensity prediction method described in this specification, in this embodiment, a comparative experiment is set up based on the actual application scenario of the typhoon intensity prediction method. Among them, the control group uses the conventional prediction method, while the verification group applies the typhoon intensity prediction method described in the present invention. Through this setting in this embodiment, the prediction results of the control group and the verification group are obtained, and a comparative analysis is carried out with the actual truth values of each typhoon. Finally, the results are as follows Figures 3 - 6The results shown. It can be understood that the mean sea level pressure, as one of the important indicators for measuring the intensity of typhoons, the improvement of its prediction accuracy is of great significance for typhoon early warning, disaster prevention and mitigation. As can be seen from the figure, the difference between the predicted mean sea level pressure of the verification group and the actual true value is significantly smaller than the difference between the predicted mean sea level pressure of the control group and the actual true value. This result indicates that the typhoon intensity prediction method described in the present invention has significant advantages in improving prediction accuracy. It is particularly worth mentioning that although the predicted mean sea level pressure of the verification group is slightly lower than the actual true value, this exactly reflects the early warning characteristics of the prediction method of the present invention. Because the decrease in mean sea level pressure often indicates the weakening of atmospheric pressure, which is usually closely related to the increased instability of weather systems, the increased possibility of storms or cyclone activities. Therefore, this small deviation in the prediction results of the verification group actually provides a more advanced and sensitive warning signal for typhoon early warning and disaster prevention and mitigation work, helping relevant departments to identify and respond to potential typhoon threats earlier.
[0065] Based on the above, this embodiment not only verifies the effectiveness of a typhoon intensity prediction method in this specification, but also further highlights its significant advantages in improving prediction accuracy and enhancing early warning capabilities, which is of great significance for improving the accuracy and timeliness of typhoon early warning systems.
[0066] Those of ordinary skill in the art can understand that all or part of the processes in implementing the method of the above-mentioned first embodiment can be completed by instructing relevant hardware through a computer program. This program can be stored in a computer-readable storage medium. When this program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: various media such as ROM, RAM, magnetic disks or optical discs that can store program codes. Without conflict, the technical features in this embodiment and the implementation scheme can be combined arbitrarily.
[0067] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.
[0068] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0069] The above are only exemplary embodiments of the present invention and should not be used to limit the scope of the present invention. That is, any equivalent changes and modifications made in accordance with the teachings of the present invention still fall within the scope covered by the present invention. Those skilled in the art will readily think of other embodiments of the present invention after considering the specification and practicing the disclosure herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common general knowledge or conventional technical means in the technical field not recorded in the present invention. The specification and examples are only regarded as exemplary, and the scope and spirit of the present invention are defined by the claims.
Claims
1. A typhoon intensity prediction method, characterized in that, Including the steps: S1. Obtain the initial typhoon field data, where the initial typhoon field data includes ERA5 reanalysis data at at least two prior times; S2. Obtain the Fuxi meteorological large model and input the initial typhoon field data thereto to output the initial typhoon prediction data at a subsequent time, where the initial typhoon prediction data includes the typhoon center position and typhoon characterization variables, and the typhoon characterization variables include basic field variables, thermal field variables, air pressure, and precipitation; S3. Calculate other characterization variables based on the output typhoon characterization variables, where the other characterization variables include diagnostic field variables and wind speed, and the basic field variables, thermal field variables, air pressure, precipitation, diagnostic field variables, and wind speed are collectively referred to as channel dimension data; S4. Select the channel dimension data within a preset diameter range based on the typhoon center position to obtain spatial dimension data; S5. Obtain preset timing parameters, arrange the spatial dimension data in chronological order to form time dimension data, and input the time dimension data into the obtained systematic deviation correction model to output the corrected typhoon intensity prediction data at a future time; The systematic deviation correction model adopts a hybrid model of a 3D convolutional neural network and a long short-term memory network. Step S5 is specifically: input the spatial dimension data corresponding to the corresponding time step into the input layer of the 3D convolutional neural network to extract the dynamic field features, thermal field features, and ground field features; Obtain the path prediction data and intensity prediction data output by the output layer of the 3D convolutional neural network; Obtain a preset loss function and correct the path prediction data and the intensity prediction data respectively to obtain the typhoon intensity prediction data for the next 24 hours and / or 48 hours and / or 72 hours.
2. A typhoon intensity prediction method according to claim 1, wherein: The basic field variables include wind field components, wind field components and geopotential height; The thermal field variables include temperature and relative humidity; The diagnostic field variables include vorticity field, divergence field, potential vorticity, and kinetic energy.
3. A typhoon intensity prediction method according to claim 2, characterized in that In step S4, the selection of the channel dimension data within a preset diameter range based on the typhoon center position to obtain spatial dimension data is specifically: Centered on the typhoon center position, select the channel dimension data with a horizontal resolution of 0.25°×0.25° and a spatial range of a diameter of 10 degrees to obtain spatial dimension data; The corresponding pixels when the spatial range is a diameter of 10 degrees are 41×41.
4. A typhoon intensity prediction method according to claim 3, wherein: The preset input timing parameter is 24 hours, the time interval is 6 hours, and there are a total of 4 time steps; The time dimension data includes each spatial dimension data corresponding to each time step.
5. A typhoon intensity prediction method according to claim 4, characterized in that, The calculation formula of the loss function is: ， In the formula, is the mean square error loss, is the mean absolute error loss, is the physical constraint loss, are the loss coefficients of the mean square error, the mean absolute error loss, and the physical constraint, respectively.
6. A typhoon intensity prediction method according to claim 5, wherein: The calculation of other characterization variables based on the typhoon characterization variables output in step S3 includes calculating the horizontal component of the divergence field, the vertical component of the vorticity field, and the kinetic energy based on the wind field component and the wind field component. The calculation formula for the horizontal component of the divergence field is , where represents the horizontal component of the divergence field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west and north-south directions respectively; The calculation formula for the vertical component of the vorticity field is , where represents the vertical component of the vorticity field, is the component of the wind field, is the component of the wind field, and are the spatial coordinates in the east-west direction and the north-south direction, respectively; The calculation formula for the kinetic energy is , where represents the kinetic energy is the component of the wind field is the component of the wind field 7. A typhoon intensity prediction system, characterized in that, For implementing the typhoon intensity prediction method according to any one of claims 1 to 6.
8. A computer device, the computer device comprising a memory, a processor, and a computer program, characterized in that, When the computer program is executed by a processor, it implements the typhoon intensity prediction method according to any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the typhoon intensity prediction method according to any one of claims 1 to 6.