A typhoon vortex initialization method and system

Abstract

The present invention discloses a typhoon vortex initialization method and system. The method includes: performing numerical simulation based on satellite observation precipitation data in the typhoon area to establish a precipitation-latent heat profile table; obtaining the latent heat distribution of different cloud types according to the precipitation-latent heat profile table; using the latent heat distribution of different cloud types as thermal forcing to drive the PV-ω equation to invert the secondary circulation at the initial moment; using the typhoon center distance as a weight coefficient to fuse the inverted secondary circulation at the initial moment with the reanalysis data environmental wind field to obtain a new initial dynamic field; performing dynamic initialization on the new initial dynamic field by using the relaxation approximation method to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, which is used as the initial typhoon vortex for subsequent forecasting. The present invention uses precipitation data to constrain the vortex dynamic initialization, thereby improving the initial vortex structure, enhancing the accuracy of describing actual typhoons, and further improving the accuracy of typhoon intensity forecasting.

Description

Technical Field

[0001] The present invention belongs to the technical field of meteorological forecasting, and particularly relates to a typhoon vortex initialization method and system. Background Art

[0002] Disastrous weather such as strong winds, heavy rains, and storm surges brought by typhoon activities often causes serious economic losses and casualties. The forecasting of typhoon paths and intensities has always been a hot topic in meteorological research. In recent years, with the improvement of technologies such as numerical models and data assimilation, the forecasting accuracy of typhoon paths has been gradually improved. However, the progress of intensity forecasting is relatively slow, and the shorter the forecasting time limit, the smaller the progress. This may be due to the defects in the physical processes of the model and the lack of observational data, as well as the insufficient model resolution, resulting in the inability to accurately describe the typhoon structure.

[0003] The initial typhoon vortex in the global model is often too weak, which requires typhoon vortex initialization when carrying out numerical forecasting, so that the intensity and structure of the initial vortex approach those of the actual typhoon. At present, all widely used typhoon vortex initialization methods have problems in accurately describing the intensity and structure of the actual typhoon, thus making it difficult to accurately forecast typhoon intensity. Summary of the Invention

[0004] To solve the deficiencies in the prior art, the present invention provides a typhoon vortex initialization method and system, which uses precipitation data to constrain the vortex dynamic initialization, thereby improving the initial vortex structure, enhancing the accuracy of describing the actual typhoon, and further improving the accuracy of typhoon intensity forecasting.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] In the first aspect, a typhoon vortex initialization method is provided, including: obtaining satellite-observed precipitation data in the typhoon area and performing preprocessing to obtain preprocessed data; performing numerical simulation based on the preprocessed data to establish a precipitation-latent heat profile table; obtaining the latent heat distribution of different cloud types in the typhoon area according to the established precipitation-latent heat profile table; using the latent heat distribution of different cloud types in the typhoon area as a thermal forcing to drive the PV-ω equation to invert the secondary circulation at the initial time; using the typhoon center distance as a weight coefficient to fuse the inverted secondary circulation at the initial time with the reanalysis data environmental wind field to obtain a new initial dynamic field; using the relaxation approximation method to perform dynamic initialization on the new initial dynamic field to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, which is used as the initial typhoon vortex for subsequent forecasting.

[0007] Further, obtaining satellite-observed precipitation data in the typhoon area and performing preprocessing includes: collecting and analyzing satellite-observed precipitation data in the typhoon area, and classifying the precipitation according to the precipitation rate to obtain preprocessed data.

[0008] Further, based on the preprocessed data, numerical simulation is carried out to establish a precipitation - latent heat profile table, including: carrying out numerical simulation using the WRF - ARW mesoscale numerical model; identifying cloud types in the typhoon area through cloud separation technology; according to the cloud types and precipitation rates of each grid point in the typhoon area in the simulation results of the numerical simulation, respectively synthesizing the average latent heat profiles corresponding to each level of precipitation rate for different cloud types, and finally establishing a precipitation - latent heat profile table.

[0009] Further, identifying cloud types in the typhoon area through cloud separation technology includes: using the total water path size to judge whether there are clouds, and identifying it as a cloudy area when the total water path is greater than the first threshold; in the cloudy area, using the ratio of the ice water path to the liquid water path to identify cloud types, and identifying it as convective cloud when the ratio is less than the second threshold, and identifying it as stratiform cloud when the ratio is greater than the second threshold.

[0010] Further, using the latent heat distribution of different cloud types in the area where the typhoon is located as the thermal forcing, driving the PV - ω equation to invert the secondary circulation at the initial moment, including: First, separating the perturbed potential vorticity field from the given initial field and inverting the balanced flow field; Second, iteratively solving the system of equations containing the quasi - balanced ω equation to obtain the non - balanced secondary circulation of the external source forcing. The specific formulas of this system of equations are as follows:

[0011] (1)

[0012] (2)

[0013] (3)

[0014] Among them, , ; and respectively represent the balanced stream function and geopotential height; , and respectively represent the velocity potential function, absolute vorticity and vertical velocity, and respectively represent the zonal and meridional surface frictions; represents the pseudo - equivalent height coordinate, , represents the scale height, represents the basic state potential temperature, represents the surface pressure, represents the diabatic heating term; represents the horizontal direction; represents the horizontal wind; represents the rotational wind; Indicates the divergent wind; Indicates the specific heat at constant volume; Indicates the specific heat at constant pressure; Indicates the gas constant of dry air; μ indicates the adiabatic index; Indicates the variation of the Coriolis parameter with latitude; Indicates the acceleration due to gravity; Indicates the Coriolis parameter; Equation (1) is the quasi-equilibrium ω equation. The forcing terms on the right side of the equation are, from left to right, the vorticity advection difference caused by the non-divergent wind and the divergent wind, the Laplacian term of the temperature advection, the differential deformation or Jacobian term, the β effect term, the non-adiabatic heating, and the friction effect term. Among them, the first four terms are dynamic forcing terms, and the last three terms are thermodynamic forcing terms; Equation (2) is used to solve the time-varying term of the stream function, and Equation (3) is used to solve the velocity potential function and the divergent wind; According to Equation (1), the latent heat heating is used as the non-adiabatic heating term, and the 3D secondary circulation structure of the typhoon is obtained by inverting the driving equation.

[0015] Furthermore, using the typhoon center distance as the weight coefficient, the secondary circulation at the initial time obtained by inversion is fused with the reanalysis data environmental wind field to obtain a new initial dynamic field, including: separating the typhoon vortex and the environmental field where the typhoon vortex is located using the spatial filtering technique; inserting the newly obtained typhoon vortex separated after inverting the PV-ω equation into the environmental field provided by the reanalysis data using the boundary smoothing technique to obtain a new initial dynamic field.

[0016] Furthermore, inserting into the environmental field provided by the reanalysis data using the boundary smoothing technique includes: during the insertion process: for the area within t1 km from the typhoon center, the vortex obtained by inversion is completely used; the area from t1 - t2 km is a linear smoothing transition zone weighted by distance; and the area outside t2 km completely uses the environmental field provided by the reanalysis data.

[0017] Furthermore, using the relaxation approximation method to perform dynamic initialization on the new initial dynamic field to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, including: adding a Newton relaxation term to the PV-ω equation, so that the set physical quantity continuously approaches the analysis field without destroying the model balance, while releasing the integration of other physical quantities within the model framework; setting the analysis field as the newly obtained initial dynamic field and performing relaxation approximation on the wind field to fix the dynamic field while releasing the integration of the thermodynamic field.

[0018] Furthermore, the relaxation approximation is implemented through the FDDA module in the ARW-WRF model, with the duration set to 12 h and the relaxation coefficient to 0.0003 s -1 .

[0019] In a second aspect, a typhoon vortex initialization system is provided, including a storage medium and a processor; the storage medium is used for storing instructions; the processor is configured to operate according to the instructions to execute the typhoon vortex initialization method described in the first aspect.

[0020] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0021] (1) The present invention uses satellite-observed precipitation data to invert the latent heat distribution, thereby constraining the typhoon initial vortex structure, which provides a more reliable data basis for typhoon vortex initialization; the observation data obtained by airborne radar and dropsondes are difficult to obtain in the northwestern Pacific Ocean, while the precipitation radar carried by satellites can give the three-dimensional structure of latent heat, but the scanning range is narrow and it is difficult to completely cover the typhoon area; in contrast, GPM-IMERG data integrates multi-source satellite scanning data in a flexible framework, has a wide coverage range, is less restricted by satellite scanning channels and is easy to obtain in real time, and can describe the typhoon structure more comprehensively.

[0022] (2) The present invention differentiates convective clouds and stratiform clouds through cloud separation technology, and gives corresponding latent heat heating profiles for different cloud types respectively; different clouds have different physical properties and vertical structures. The latent heat heating of convective clouds is concentrated at relatively low altitudes, while stratiform clouds often correspond to cooling in the lower layer and the latent heat heating is located at higher altitudes; the cloud separation method adopted by the present invention can effectively separate the two types of clouds, and the finally obtained latent heat heating profile is more in line with the actual situation.

[0023] (3) The present invention uses the PV-ω equation to invert the secondary circulation structure of the typhoon, improving the asymmetric distribution of the typhoon boundary layer inflow and upper-level outflow structures; the secondary circulation is an important part of the typhoon. The inflow in the lower troposphere causes the convergence of water vapor and angular momentum, while the upper-level outflow transports the angular momentum and mass in the inner area of the typhoon outward, thus promoting the development of the system. There is often an obvious asymmetric structure in the typhoon outflow channel, and it has an important connection with the typhoon intensity change; compared with the two-dimensional Sawyer-Eliassen (SE) equation used in traditional diagnostic analysis, the PV-ω equation can obtain the three-dimensional asymmetric structure of the secondary circulation, and a more accurate initial vortex secondary circulation structure can effectively improve the subsequent typhoon intensity prediction quality.

[0024] (4) The present invention combines the relaxation approximation and the dynamic initialization technology to obtain a thermal field that is in balance with the dynamic field; there is a close relationship between the typhoon intensity and the warm core intensity and height. If only the wind field is improved without improving the pressure and temperature fields, it may lead to the imbalance of the initial field; the relaxation approximation adopted by the present invention can obtain an initial vortex with a thermodynamically balanced structure under the constraint of the model physical framework, and consumes less computing resources, largely avoiding the model spin down caused by imbalance. Description of the Drawings

[0025] Figure 1 is a schematic diagram of the implementation process of a typhoon vortex initialization method provided by an embodiment of the present invention;

[0026] Figure 2 is the horizontal distribution of latent heat at the 600 hPa height obtained by looking up the table three hours before the initial moment;

[0027] Figure 3 is the latent heat distribution at a height of 15 km in the outflow layer and the divergence wind field inverted by the PV - ω equation;

[0028] Figure 4 is a comparison chart of the typhoon intensity and the best - track data predicted before and after using this initialization method. Among them, (a) is the comparison of the evolution of the maximum wind speed at the typhoon center, and (b) is the comparison of the evolution of the minimum sea - level pressure at the typhoon center;

[0029] Figure 5 is a comparison chart of the typhoon radial wind structure at the initial moment before and after using this initialization method. Among them, (a) is the typhoon radial wind structure without initialization, and (b) is the typhoon radial wind structure initialized by the method of the present invention;

[0030] Figure 6 is a comparison chart of the typhoon warm - core structure at the initial moment before and after using this initialization method. Among them, (a) is the typhoon warm - core structure without initialization, and (b) is the typhoon warm - core structure initialized by the method of the present invention. Detailed Implementation Manner

[0031] The present invention will be further described below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and should not be used to limit the protection scope of the present invention.

[0032] Embodiment 1

[0033] As Figures 1 to 6As shown in the figure, a typhoon vortex initialization method based on satellite observation and the PV-ω equation includes: obtaining satellite observation precipitation data in the typhoon area and performing preprocessing to obtain preprocessed data; performing numerical simulation based on the preprocessed data to establish a precipitation-latent heat profile table; obtaining the latent heat distribution of different cloud types in the area where the typhoon is located according to the established precipitation-latent heat profile table; using the latent heat distribution of different cloud types in the area where the typhoon is located as a thermal forcing to drive the PV-ω equation to invert the secondary circulation at the initial time; using the typhoon center distance as a weight coefficient to fuse the inverted secondary circulation at the initial time with the reanalysis data environmental wind field to obtain a new initial dynamic field; using the relaxation approximation method to perform dynamic initialization on the new initial dynamic field to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, which is used as the initial typhoon vortex for subsequent forecasting.

[0034] To implement the typhoon vortex dynamic initialization method based on satellite observation and the PV-ω equation, the Advanced Research WRF (ARW) (version 3.9.1) in the Weather Research and Forecasting Model (WRF) is used, and the typhoon "Lekima" (2019) is selected as an example to further illustrate the method of the present invention.

[0035] The model uses a triple-nested grid with horizontal resolutions of 18, 6, and 2 km respectively, and 45 layers in the vertical direction. The initial field and lateral boundary conditions come from the National Centers for Environmental Prediction / Final Operational Global Analysis (NCEP / FNL). The initial time of the experiment is 18:00 on August 6, 2019, and the simulation time is 48 hours. As Figure 1 shown, the typhoon vortex initialization method based on satellite observation and the PV-ω equation of the present invention mainly includes the following steps.

[0036] S1. Obtain satellite observation precipitation data in the typhoon area and perform preprocessing to obtain preprocessed data.

[0037] Collect and analyze the final operational version of the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (GPM-IMERG Final Run). Classify precipitation according to the precipitation rate, with intervals of 5 mm / h from 0 to 45 mm / h and one interval for precipitation above 45 mm / h, for a total of 10 intervals, as preprocessed data.

[0038] S2. Conduct numerical simulations based on the preprocessed data to establish a precipitation-latent heat profile table.

[0039] Use the WRF-ARW mesoscale numerical model for long-term numerical simulations. Distinguish convective cloud regions and stratiform cloud regions through cloud separation techniques. According to the cloud types and precipitation rates at each grid point in the area near the typhoon in the simulation results, synthesize the average latent heat profiles corresponding to each precipitation rate interval for the convective cloud region and the stratiform cloud region respectively, and finally establish a precipitation-latent heat profile table.

[0040] The cloud separation method is as follows: Distinguish cloud types based on the content of various hydrometeors in the grid air column. Use the total water path (the column-integrated mixing ratio of all hydrometeors) to judge whether there are clouds. When the total water path is greater than 0.1 mm (the first threshold), it is identified as a cloudy area; use the ratio of the ice water path (the column-integrated mixing ratio of ice-phase hydrometeors such as graupel, snow, and ice) to the liquid water path (the column-integrated mixing ratio of liquid-phase hydrometeors such as rain and cloud liquid water) to identify cloud types. When the ratio is less than 2 (the second threshold), it is identified as a convective cloud, and when the ratio is greater than 2, it is identified as a stratiform cloud.

[0041] S3. Obtain the latent heat distribution of different cloud types in the area where the typhoon is located according to the established precipitation-latent heat profile table.

[0042] Select the GPM-IMERG precipitation rate distribution 3 hours before the initial time. Distinguish the cloud types at each grid point in the area near the typhoon through the Nanjing Joint Institute for Atmospheric Sciences Himawari-8 / 9 Cloud Feature Dataset (NJIAS_HCFD), and give the latent heat distribution at each grid point by looking up the table, as Figure 2 shown.

[0043] S4. Use the latent heat distribution of different cloud types in the area where the typhoon is located as the thermal forcing to drive the PV-ω equation to invert the secondary circulation at the initial time.

[0044] The PV-ω equation inversion involves: First, separating the perturbation potential vorticity (PV) field from the given initial field and obtaining the balanced flow field through PV inversion; Second, iteratively solving the system of equations containing the quasi-balanced ω equation to obtain the non-balanced secondary circulation of the external source forcing. The specific formulas of this system of equations are as follows:

[0045] (1)

[0046] (2)

[0047] (3)

[0048] Among them, , ; and represent the balanced stream function and geopotential height respectively; , and represent the velocity potential function, absolute vorticity and vertical velocity respectively, and represent the zonal and meridional surface frictions respectively; represents the pseudo-equivalent height coordinate, , represents the scale height, represents the basic state potential temperature, represents the surface pressure, represents the diabatic heating term; represents the horizontal direction; represents the horizontal wind; represents the rotational wind; represents the divergent wind; represents the specific heat at constant volume; represents the specific heat at constant pressure; represents the dry air gas constant; μ represents the adiabatic index; represents the variation of the Coriolis parameter with latitude; represents the acceleration due to gravity; represents the Coriolis parameter.

[0049] The quasi-balanced ω equation is formula (1), where the forcing terms on the right side of the equal sign are, from left to right, the vorticity advection difference caused by the non-divergent wind and the divergent wind, the Laplacian term of the temperature advection, the differential deformation or Jacobian term, the β effect term, the diabatic heating, and the friction effect term; among them, the first four terms are dynamic forcing terms, and the last three terms are thermodynamic forcing terms; formula (2) is used to solve the term of the stream function varying with time, and formula (3) is used to solve the velocity potential function and the divergent wind; according to formula (1), it can be known that: the thermodynamic forcing mainly depends on the diabatic heating. Therefore, the latent heat heating is used as the diabatic heating term, and the 3D secondary circulation structure (divergent wind and vertical motion) of the typhoon is obtained by inverting the driving equation.

[0050] The divergent wind field of the typhoon outflow is largely affected by the convection distribution. When the typhoon convection is stronger, the outflow divergent wind is also stronger. Asymmetric convection can affect the distribution of the typhoon outflow channels by modulating the intensity of the outflow divergent wind. As Figure 3 shown, the relatively strong latent heat distribution on the southwest side of the typhoon forces out a relatively strong typhoon outflow divergent wind, while the relatively weak latent heat on the north side forces out a weaker divergent wind, and the typhoon outflow shows an obvious asymmetric structure. The outflow layer in the typhoon system is manifested as the high-level air flow outflow interval. On the synoptic scale, the outflow layer is a shallow anticyclone circulation system with divergence characteristics, generally located at an altitude of 100 - 300 hPa.

[0051] S5. Taking the typhoon center distance as the weight coefficient, the secondary circulation at the initial moment obtained by inversion is fused with the environmental wind field of the NCEP / FNL reanalysis data to obtain a new initial dynamic field.

[0052] The wind field fusion method is as follows: The Tukey window function (Tukey-Window) spatial filtering technology is used to separate the vortex and the environmental field. The separation cut-off wavelength is 500 km. The disturbance system smaller than 500 km is regarded as the typhoon vortex, and the part larger than 500 km is regarded as the environmental field. After the inversion of the PV-ω equation, the newly obtained typhoon vortex is separated and inserted into the environmental field provided by the NECP / FNL reanalysis data. During the insertion process, a boundary smoothing technology is adopted: for the area within t1 = 300 km from the typhoon center, the vortex obtained by inversion is completely adopted, for the area outside t2 = 500 km, the environmental field provided by the reanalysis data is completely adopted, and the area between 300 km and 500 km is a linear smoothing transition zone weighted by the distance.

[0053] S6. The new initial dynamic field is dynamically initialized using the relaxation approximation method to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, which is used as the initial typhoon vortex for subsequent forecasts.

[0054] Dynamic initialization is carried out using the relaxation approximation method. The dynamic fields (U; V wind fields) are fixed, and other physical quantities are updated through model integration, so that the thermodynamic fields (pressure and temperature fields) are adjusted towards the dynamic fields (wind fields). Finally, a vortex in thermal-dynamic balance under the constraint of the model physical framework is obtained as the initial typhoon vortex for subsequent forecasting.

[0055] The relaxation approximation method is as follows: By adding an appropriate Newton relaxation term to the control equation, some physical quantities can continuously approach the analysis field without destroying the model balance, while other physical quantities are integrated within the model framework. The analysis field is set as the initial field obtained in S5, and relaxation approximation is performed on the wind field, so that the dynamic fields are fixed while the thermodynamic fields are integrated freely. The relaxation approximation is realized through the Four Dimensional Data Assimilation (FDDA) module in the ARW-WRF model. The duration is set to 12 h, and the relaxation coefficient is 0.0003 s -1 。

[0056] Figure 4 The forecast results in [reference] show that the typhoon undergoes rapid intensification within 24 hours after the start of the simulation. The evolutions of the maximum wind speed and the minimum pressure are both close to the observed results given by the best track. Without initialization, the typhoon intensity increases slowly. At the 24th hour (18:00 on the 7th) of the simulation, the maximum wind speed only reaches 80 kts (80 knots), and the rapid intensification process cannot be accurately predicted. After initialization, the structure of the initial typhoon vortex has been significantly improved. The wind field inverted by the PV-ω equation has stronger upper-level outflows compared with the reanalysis data, as shown in Figure 5 shown. The relaxation approximation method forces the warm core of the typhoon to intensify (as shown in Figure 6 shown) and the central pressure to decrease, reducing the imbalance between the thermal and dynamic fields.

[0057] The present invention uses satellite-observed precipitation data to constrain the asymmetric structure of the initial vortex. Its spatial distribution has a good correspondence with the typhoon structure, and it has a wide coverage range and is easy to obtain in real time. The method of the present invention constructs the initial typhoon vortex through the PV-ω equation inversion and relaxation approximation technology, with less computational resource consumption and the obtained initial vortex in thermal-dynamic balance. The present invention effectively integrates and utilizes observation data such as precipitation to constrain the structure of the initial typhoon vortex and improves the typhoon intensity forecast quality.

[0058] Example 2

[0059] Based on the typhoon vortex initialization method based on satellite observation and PV-ω equation described in Embodiment 1, this embodiment provides a typhoon vortex initialization system based on satellite observation and PV-ω equation, including a storage medium and a processor; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the typhoon vortex initialization method based on satellite observation and PV-ω equation described in Embodiment 1.

[0060] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A typhoon vortex initialization method, characterized in that Including: Obtain satellite observation precipitation data of the typhoon area and perform preprocessing to obtain preprocessed data; Based on the preprocessed data, conduct numerical simulation to establish a precipitation-latent heat profile table; Obtain the latent heat distribution of different cloud types within the typhoon area according to the established precipitation-latent heat profile table; Use the latent heat distribution of different cloud types within the typhoon area as a thermal forcing to drive the inversion of the PV-ω equation to obtain the secondary circulation at the initial time; Use the typhoon center distance as a weight coefficient to fuse the inverted secondary circulation at the initial time with the reanalysis data environmental wind field to obtain a new initial dynamic field; Use the relaxation approximation method to perform dynamic initialization on the new initial dynamic field to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, which is used as the initial typhoon vortex for subsequent forecasts; Among them, obtaining satellite observation precipitation data of the typhoon area and performing preprocessing includes: collecting and analyzing satellite observation precipitation data of the typhoon area, dividing the precipitation according to the precipitation rate to obtain preprocessed data; Based on the preprocessed data, conduct numerical simulation to establish a precipitation-latent heat profile table, including: Use the WRF-ARW mesoscale numerical model to conduct numerical simulation; Identify the cloud types in the typhoon area through cloud separation technology; According to the cloud types and precipitation rate magnitudes of each grid point in the typhoon area in the simulation results of the numerical simulation, respectively synthesize the average latent heat profiles corresponding to each precipitation rate of different cloud types, and finally establish a precipitation-latent heat profile table; Identifying the cloud types in the typhoon area through cloud separation technology includes: Use the total water path magnitude to judge whether there are clouds. When the total water path is greater than the first threshold, it is identified as a cloudy area; In the cloudy area, use the ratio of the ice water path to the liquid water path to identify the cloud type. When the ratio is less than the second threshold, it is identified as convective cloud, and when the ratio is greater than the second threshold, it is identified as stratiform cloud. Identifying the cloud types in the typhoon area through cloud separation technology includes: Use the total water path magnitude to judge whether there are clouds. When the total water path is greater than the first threshold, it is identified as a cloudy area; In the cloudy area, use the ratio of the ice water path to the liquid water path to identify the cloud type. When the ratio is less than the second threshold, it is identified as convective cloud, and when the ratio is greater than the second threshold, it is identified as stratiform cloud.

2. The typhoon vortex initialization method according to claim 1, wherein, Using the latent heat distribution of different cloud types within the typhoon area as a thermal forcing to drive the inversion of the PV-ω equation to obtain the secondary circulation at the initial time, including: First, separate the perturbed potential vorticity field from the given initial field and invert the balanced flow field; second, iteratively solve the system of equations containing the quasi-balanced ω equation to obtain the non-equilibrium secondary circulation of the external source forcing. The specific formula of this system of equations is shown as follows: （1） （2） （3） Among them, , ; and respectively represent the balanced stream function and the geopotential height; , and respectively represent the velocity potential function, the absolute vorticity and the vertical velocity, and respectively represent the zonal x and the meridional y surface friction; represents the pseudo-equivalent height coordinate, , represents the scale height, represents the basic state potential temperature, represents the surface pressure, represents the diabatic heating term; represents the horizontal direction; represents the horizontal wind; represents the rotational wind; represents the divergent wind; represents the specific heat at constant volume; represents the specific heat at constant pressure; represents the dry air gas constant; μ represents the adiabatic index; represents the variation of the Coriolis parameter with latitude; represents the acceleration due to gravity; represents the Coriolis parameter; The formula (1) is the quasi-balanced ω equation. The forcing terms on the right side of the equal sign are, from left to right, the vorticity advection difference caused by the non-divergent wind and the divergent wind, the Laplacian term of the temperature advection, the differential deformation or Jacobian term, the β effect term, the diabatic heating, and the friction effect term. Among them, the first four terms are dynamic forcing terms, and the last three terms are thermodynamic forcing terms. The formula (2) is used to solve the time-varying term of the stream function, and the formula (3) is used to solve the velocity potential function and the divergent wind. According to the formula (1), the latent heat heating is used as the diabatic heating term, and the 3D secondary circulation structure of the typhoon is obtained by inverting the driving equation.

3. The typhoon vortex initialization method according to claim 2, characterized in that Taking the typhoon center distance as the weight coefficient, the secondary circulation at the initial moment obtained by inversion is integrated with the reanalysis data environmental wind field to obtain a new initial dynamic field, including: Using the spatial filtering technique to separate the typhoon vortex and the environmental field where the typhoon vortex is located; The newly obtained typhoon vortex separated after inverting the PV-ω equation is inserted into the environmental field provided by the reanalysis data using the boundary smoothing technique to obtain a new initial dynamic field.

4. The typhoon vortex initialization method according to claim 3, wherein, Inserted into the environmental field provided by the reanalysis data using the boundary smoothing technique, including: During the insertion process: for the area within t1 km from the typhoon center, the vortex obtained by inversion is completely used; the area from t1 - t2 km is a linear smoothing transition zone weighted by distance; and the area outside t2 km completely uses the environmental field provided by the reanalysis data.

5. The typhoon vortex initialization method according to claim 3, wherein Using the relaxation approximation method to perform dynamic initialization on the new initial dynamic field to obtain a vortex in thermal dynamic equilibrium under the constraint of the model physical framework, including: By adding a Newton relaxation term to the PV-ω equation, the set physical quantity continuously approaches the analysis field without destroying the model balance, while other physical quantities are integrated under the model framework; the analysis field is set as the newly obtained initial dynamic field, and the wind field is relaxed and approximated to fix the dynamic field while the thermodynamic field is integrated freely.

6. The typhoon vortex initialization method according to claim 5, wherein The relaxation approximation is achieved through the FDDA module in the ARW-WRF model, with a duration set to 12 h and a relaxation coefficient of 0.0003 s -1 .

7. A typhoon vortex initialization system, characterized in that, Including a storage medium and a processor; The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the typhoon vortex initialization method described in any one of claims 1 to 6.

Images