Rapid enhanced tropical cyclone precipitation structure parting and environmental influence assessment method

Abstract

The invention relates to the technical field of atmospheric science, and discloses a rapid enhanced tropical cyclone precipitation structure parting and environmental influence assessment method. According to the method, a self-organizing mapping technology is utilized to perform clustering analysis on precipitation distribution images of the fast enhanced tropical cyclones, and possible precipitation structure types of the fast enhanced tropical cyclones are clarified. Then, the convection and dynamic characteristics of various types of tropical cyclones are quantitatively tested from the aspects of convection intensity, symmetry, vortex scale, gradient and the like. And finally, comprehensively evaluating the influence of a plurality of large-scale environmental factors such as vertical wind shear, sea surface temperature and latent heat flux on the tropical cyclone precipitation structure through synthetic analysis and correlation analysis methods, and clarifying the physical cause of the difference of the tropical cyclone precipitation structure. The method has the beneficial effects that the method is beneficial to quantitatively evaluating the difference and diversity of the rapid enhancement type tropical cyclone precipitation structure and dynamic characteristics, reveals the internal relation between rapid enhancement and a large-scale environment, and provides an important scientific basis for improving the tropical cyclone forecasting accuracy and resisting tropical cyclone disasters.

Description

Technical Field

[0001] This invention relates to the field of atmospheric science and technology, and in particular to a method for classifying the precipitation structure of rapidly intensifying tropical cyclones and assessing their environmental impact. Background Technology

[0002] Tropical cyclones are among the most destructive weather systems affecting coastal areas. Their rapid intensification is often sudden and intense, posing significant challenges to disaster prevention and mitigation. Rapid intensification refers to the dramatic increase in wind speed within a short period. Existing numerical models and forecasting methods struggle to capture this process accurately and promptly, leading to substantial discrepancies between forecasts and actual conditions. Currently, research and forecasting of rapid intensification primarily rely on the analysis of large-scale environmental factors, such as sea surface temperature, humidity, and vertical wind shear. However, these factors exhibit complex nonlinear relationships with the internal precipitation structure and convection organization of cyclones, making it difficult to accurately reflect the mechanisms underlying rapid intensification using a single indicator.

[0003] Observational and simulation results show significant differences in precipitation distribution and convective patterns among rapidly intensifying tropical cyclones. Traditional methods, often based on symmetry indices or empirical statistics, struggle to accurately distinguish different precipitation structure types and fail to reveal their intrinsic coupling relationships with environmental factors, thus limiting the improvement of rapid intensification mechanism research and forecasting capabilities. To address these issues, this invention proposes a method for categorizing precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones. This method utilizes self-organizing map clustering to objectively identify precipitation structure patterns and combines this with quantitative analysis of features such as convective intensity and convective symmetry to reveal the mechanisms by which environmental factors influence different structural patterns. This invention effectively overcomes the shortcomings of existing technologies, providing crucial support for improving the accuracy of rapid intensification forecasts and typhoon disaster prevention. Summary of the Invention

[0004] The purpose of this method is to address the shortcomings of existing research methods on the structure of rapidly intensifying tropical cyclone precipitation and its relationship with large-scale environments by proposing a new method for classifying the structure of rapidly intensifying tropical cyclone precipitation and assessing its environmental impact.

[0005] The rapid enhanced tropical cyclone precipitation structure classification and environmental impact assessment method of the present invention includes the following steps: Step S1: Data Acquisition; Acquire measured optimal tropical cyclone track data, global satellite precipitation data, and meteorological reanalysis data; Step S2: Screening of rapidly intensifying tropical cyclones; Using the optimal track data of tropical cyclones obtained in Step S1, screen for rapidly intensifying tropical cyclones that meet the criteria based on the objective definition of rapidly intensifying events, and extract their intensity and location information. Step S3: Extraction and preprocessing of tropical cyclone satellite precipitation images; Combining the location information of rapidly intensifying tropical cyclones obtained in Step S2, precipitation images at the moment of rapid intensification of tropical cyclones are extracted from global satellite precipitation data; Subsequently, based on the meteorological reanalysis data obtained in Step S1, the environmental vertical wind shear experienced by each tropical cyclone case is calculated, and the precipitation images are rotated according to the direction of vertical wind shear to unify them into a precipitation distribution field under a reference system with the wind shear pointing due north; Step S4: Self-organizing map clustering of precipitation images; Self-organizing map neural network clustering technology is used to objectively classify the precipitation images of rapidly intensifying tropical cyclones obtained in Step S3 to clarify the possible precipitation structure types of rapidly intensifying tropical cyclones. Step S5: Analysis of convective and dynamic characteristics of individual tropical cyclones of different types; Based on the tropical cyclone clustering results of Step S4, combined with satellite precipitation data and optimal tropical cyclone track data, the convective and dynamic characteristics of tropical cyclones in different clusters are statistically analyzed, and the differences between different tropical cyclone groups are compared. Step S6: Large-scale environmental factor preprocessing; Based on the meteorological reanalysis data obtained in Step S1, directly extract or indirectly calculate the large-scale environmental variables near individual tropical cyclones, and based on the method in Step S3, rotate the large-scale environmental variable field to unify it into a variable distribution field under a reference system with the wind shear pointing due north. Step S7: Correlation analysis of the causes of differences in tropical cyclone precipitation structure; combined with the large-scale environmental variable field obtained in Step S6, calculate the mean values ​​of environmental factors in the activity area of ​​rapidly intensifying tropical cyclones, and then use the Pearson correlation coefficient to calculate the correlation between each environmental factor and the tropical cyclone convection characteristic index obtained in Step S5. Step S8: Perform large-scale environmental factor synthesis analysis on individual cases of different types of tropical cyclones; based on the clustering and typing results of step S4, synthesize the environmental fields corresponding to each type of rapidly intensifying tropical cyclone, calculate the spatial distribution of environmental field differences between different groups, and then evaluate the impact of spatial differences in environmental fields on convective structure, and clarify the causes of differences in precipitation structure of different rapidly intensifying tropical cyclones.

[0006] A storage device that stores instructions and data for implementing a rapid enhancement method for tropical cyclone precipitation structure classification and environmental impact assessment.

[0007] A device for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment includes: a processor and a storage device; the processor loads and executes instructions and data in the storage device to implement a method for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment.

[0008] The beneficial effects provided by this invention are: (1) This invention proposes an objective classification method for the precipitation structure of rapidly intensifying tropical cyclones, which can identify a variety of convection patterns that may occur during the rapid intensification process, and helps to deepen the understanding of the structural characteristics of rapidly intensifying tropical cyclones and the diversity of their physical mechanisms.

[0009] (2) This invention combines large-scale environmental analysis of different types of tropical cyclones to reveal the control effect of environmental factors on various precipitation structures and clarify the environmental impact mechanism of different rapid intensification types, providing new ideas and directions for improving rapid intensification forecast models and enhancing the accuracy of tropical cyclone intensity forecasts. Attached Figure Description

[0010] Figure 1 This is a flowchart of the implementation of the method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones; Figure 2 This is a schematic diagram showing the effect of the tropical cyclone precipitation field rotating and unifying to a reference frame where the wind shear points due north. The black arrows indicate the direction of the wind shear. Figure 3 This is the clustering and classification results of precipitation structure of rapidly intensifying tropical cyclones when the total number of clusters is 2 to 5, i.e., the composite precipitation rate map of each cluster. Figure 4 It is a box plot analysis of the convective characteristics (maximum precipitation intensity, symmetry) and dynamic characteristics (cyclone intensity, scale and tilt) of tropical cyclones in different clusters; Figure 5 Box plot analysis of regional mean values ​​of environmental factors for tropical cyclones in different clusters; Figure 6 This is a correlation analysis diagram of environmental factors and their correlation with tropical cyclone convection characteristics. Figure 7 This is the spatial distribution field of large-scale environmental factor differences between samples of rapidly intensifying tropical cyclones with inverse shear left-side (USL) and other tropical cyclones. The green diagonal line indicates that the differences are significant at the 95% confidence level. Figure 8 This is the spatial distribution field of large-scale environmental factor differences between samples of rapidly enhanced tropical cyclones with shear left-side type (DSL) and other tropical cyclones. The green diagonal line indicates that the differences are significant at the 95% confidence level. Figure 9 This is the spatial distribution field of large-scale environmental factor differences between samples of right-side shear-reinforced tropical cyclones (DSR) and other tropical cyclones. The green diagonal line indicates that the differences are significant at the 95% confidence level. Figure 10 This is a schematic diagram of the hardware device of the present invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0012] The present invention will first explain the relevant basic concepts and the core points of this application as follows, and then elaborate on the technical solution of the present invention.

[0013] Please refer to Figure 1 , Figure 1 This is a schematic diagram of the method flow of the present invention; a method for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment, comprising the following steps: Step S1: Data Acquisition; Acquire measured optimal tropical cyclone track data, global satellite precipitation data, and meteorological reanalysis data; In step S1, the optimal path data for the tropical cyclone includes the center latitude and longitude of the tropical cyclone and the maximum wind speed near the sea surface, with a time resolution of 6 hours; the global satellite precipitation data is high spatiotemporal resolution gridded precipitation data, with a time resolution of 30 minutes per hour and a spatial resolution of 0.1°×0.1°; the meteorological reanalysis data includes meridional wind, zonal wind, atmospheric temperature, relative humidity, sea surface temperature, wind speed at 10 m altitude, and specific humidity at 2 m altitude.

[0014] Step S2: Screening of rapidly intensifying tropical cyclones; Using the optimal track data of tropical cyclones obtained in S1, screen rapidly intensifying tropical cyclones that meet the conditions based on the objective definition of rapidly intensifying events, and extract their intensity and location information. In step S2, the screening criteria for rapidly intensifying tropical cyclones include three steps: S21: The rate of change of intensity of a tropical cyclone during its rapid intensification phase shall not be less than 30 knots / day; S22: The intensity of the tropical cyclone reaches the level of a severe tropical storm or above at the moment of its rapid intensification. S23: The tropical cyclone was located south of 30°N during its rapid intensification phase and did not make landfall.

[0015] Step S3: Extraction and preprocessing of tropical cyclone satellite precipitation images; Combining the latitude and longitude location information of rapidly intensifying tropical cyclones obtained in S2, precipitation images at the start of rapid intensification are extracted from global satellite precipitation data. Subsequently, the environmental vertical wind shear is calculated for each tropical cyclone case based on meteorological reanalysis data. The precipitation images are rotated according to the direction of the vertical wind shear to unify them into a precipitation distribution field under a reference frame where the wind shear points due north. In step S3, the precipitation image extracted from global satellite precipitation data at the start of rapid enhancement is centered on the cyclone center and has a range of 400 km × 400 km. The environmental vertical wind shear is defined as the difference between the average wind vector of the 200 hPa and 800 hPa pressure layers within a radius of 400 km–800 km from the typhoon center, calculated using the following formula:

[0016]

[0017] In the formula, VWS For vertical wind shear variable level, D The vertical wind shear angle is expressed in radians. , This represents the average zonal and meridional wind vectors within a radius of 400–800 km. The subscripts 200 and 850 indicate that the variable is located at a level of 200 or 850 hPa.

[0018] Step S4: Precipitation image self-organizing map clustering; The precipitation images of rapidly intensifying tropical cyclones obtained in S3 are objectively classified using the self-organizing map (Self-Organizing Map) neural network clustering technique to clarify the possible precipitation structure types of rapidly intensifying tropical cyclones. In step S4, the self-organizing map neural network is an unsupervised learning neural network whose main goal is to map high-dimensional data onto a two-dimensional regular grid while preserving the data's topological structure, so that similar data fall into adjacent or close positions after mapping. Clustering tropical precipitation images using this technique involves the following three steps: S41: Generate a random weight matrix, typically initialized within the [0,1] interval or the data range, to ensure diversity in the initial states of the neurons and avoid premature bias towards a particular pattern during training. S42: The precipitation rate in the precipitation image is normalized using the following formula:

[0019] In the formula X and X′ Distinguish between raw precipitation rate and standardized precipitation rate. n This represents the total number of grid points in the two-dimensional field of precipitation rate. This homogenization process can eliminate the differences in overall precipitation intensity among different typhoon cases, thus allowing clustering to focus more on the spatial distribution structure of precipitation. S43: Calculate the average variance ratio ( VR And determine the optimal number of clusters, the formula for which is:

[0020] in This represents the squared distance between the centroid of a given cluster and the centroid of the total sample. This represents the mean of the squared distances between all samples and the total sample in the cluster and the centroid. VR A higher mean indicates more significant differences between clusters. VR The number of clusters increases monotonically, while the magnitude of change gradually decreases. VR When the number of clusters reaches a certain critical number and then plateaus, that number is taken as the optimal number of clusters.

[0021] Step S5: Analysis of convective and dynamic characteristics of individual tropical cyclones of different types; Based on the tropical cyclone clustering results in S4, combined with satellite precipitation data and optimal tropical cyclone track data, the convective and dynamic characteristics of different clusters are statistically analyzed, and the differences between different tropical cyclone groups are compared. In step S5, statistical characteristics of convective and dynamic factors such as maximum precipitation intensity, convective symmetry, cyclone intensity, scale, and tilt are extracted for different tropical cyclone groups mainly through synthetic analysis (i.e., calculating the multi-sample mean of individual cases of various clusters of tropical cyclones). The maximum precipitation intensity is defined as the maximum precipitation rate within a 200 km radius. Convective symmetry is measured using the precipitation asymmetry index (PAI), calculated using the following formula:

[0022] in R n = R 1–6 / R 0 represents the standardized precipitation rate. R 1–6 and R 0 represents the sum of the asymmetric components and the symmetric component of the precipitation rate for waves 1–6, respectively. r and λ These represent the radius and azimuth, respectively. A larger PAI indicates a more asymmetrical precipitation structure. The intensity of a tropical cyclone is measured by the maximum near-shore wind speed, and its scale is determined by the radius of the gust circle, which is 34 knots (numerical value 17). The radius of the wind speed circle is measured, and the tilt of the vortex is measured by the relative distance between the vortex centers at 500 hPa and 850 hPa.

[0023] Step S6: Large-scale environmental factor preprocessing; Based on the meteorological reanalysis data obtained in Step S1, directly extract or indirectly calculate the large-scale environmental variables near individual tropical cyclones, and based on the method in Step S3, rotate the large-scale environmental variable field to unify it into a variable distribution field under a reference system with the wind shear pointing due north. Step S6 mainly includes the following two steps: S61: Calculate the average relative humidity of the 700–500 hPa pressure layer as an assessment standard for mid-tropospheric humidity; calculate upper-level divergence based on the meridional and zonal wind fields of the 200 hPa pressure layer; calculate the sea surface latent heat flux (LHF) based on the overall aerodynamic formula, as follows:

[0024] In the formula L v Represents the latent heat of phase change of water vapor (value is) ), C k Represents the latent heat exchange coefficient (value is) ), ρ d The density of dry air (value is) ), U 10 At a height of 10 m, q *and q 2 represents the saturated specific humidity at the sea surface and at a height of 2 m, respectively.

[0025] S62: Extract the environmental variable field of 1000km×1000km centered on the tropical cyclone, and rotate it uniformly to a reference frame in which the wind shear points due north, following the method in step S3.

[0026] Step S7: Correlation analysis of the causes of differences in tropical cyclone precipitation structure; combined with the large-scale environmental variable field obtained in Step S6, calculate the mean values ​​of environmental factors in the activity area of ​​rapidly intensifying tropical cyclones, and then use the Pearson correlation coefficient to calculate the correlation between each environmental factor and the tropical cyclone convection characteristic index obtained in Step S5. In step S7, the correlation between each environmental factor and the convective characteristic index of each tropical cyclone is evaluated by calculating the Pearson correlation coefficient, as shown in the following formula:

[0027] in X Represents any environmental factor. Y Cov represents the convective characteristics of any tropical cyclone. X , Y ) represents the covariance of two variables. σ Let E be the standard deviation and E be the expected value.

[0028] Step S8: Perform large-scale environmental factor synthesis analysis on individual cases of different types of tropical cyclones. Based on the clustering and typing results of Step S4, synthesize the environmental fields corresponding to each type of rapidly intensifying tropical cyclone, calculate the spatial distribution of environmental field differences between different groups, and then assess the impact of spatial differences in the environmental field on the convective structure to clarify the causes of differences in precipitation structure among different rapidly intensifying tropical cyclones.

[0029] In step S8, when evaluating the large-scale environmental conditions of different clusters, the horizontal space is divided into four quadrants relative to the direction of environmental wind shear: left side of the shear, right side of the shear, left side of the wind shear, and right side of the wind shear, where "towards" and "opposite" represent the direction of the wind shear and the direction of the wind shear, respectively.

[0030] The following uses tropical cyclones (TCs) that formed in the North Pacific and North Atlantic from 2000 to 2020 as specific examples for further technical description of the present invention. These examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0031] The implementation flowchart of the rapid enhanced tropical cyclone precipitation structure classification and environmental impact assessment method of this invention is as follows: Figure 1 As shown, the specific steps are as follows: (1) Observational data acquisition; The optimal track data for tropical cyclones collected in this embodiment are from the International Best Track Archive for Climate Stewardship version 4 (IBTrACS V4) dataset, including the center latitude and longitude of the tropical cyclone (TC) and the maximum sustained wind speed at 10m altitude, with a temporal resolution of 3 hours. The Integrated Multi-satellit Eretrievals for GPM (IMERG) global satellite precipitation dataset was also collected to characterize the intensity and distribution of convective activity in the TC; this dataset has a temporal resolution of 0.5 hours and a horizontal resolution of 0.1° × 0.1°. The fifth-generation reanalysis data (ERA5) from the European Centre for Medium-Range Weather Forecasts (ECMWF) was also collected, with variables including meridional wind, zonal wind, relative humidity, wind field at 10m altitude, specific humidity at 2m altitude, and sea surface temperature; the data has a horizontal resolution of 0.1° × 0.1° and a temporal resolution of 1 hour. It should be noted that the data collected in this embodiment covers the period from 2000 to 2020. Furthermore, data from each dataset was analyzed at 6-hour intervals (0, 6, 12, and 18 UTC). Table 1 shows the information from the collected data: Table 1 Main Data Information

[0032] (2) Screening of rapidly intensifying tropical cyclones; In this embodiment, based on IBTRACSV4 optimal path data, rapidly intensifying tropical cyclones that formed in the North Pacific and North Atlantic during the period of 2000–2020 were screened. The specific screening criteria for rapidly intensifying tropical cyclones are as follows: ① The intensification rate of the tropical cyclone during the rapid intensification phase is not less than 30 knots / day (1 knot = 0.51 knots / day). ); ② Rapidly intensify to the level of a severe tropical storm or above at the time of initiation (maximum sustained wind speed at 10 m height greater than or equal to 17). ); ③ The tropical cyclone remained south of 30°N latitude throughout its rapid intensification phase and did not make landfall within the next 24 hours. Based on these four conditions, a total of 434 cases of rapidly intensifying tropical cyclones were selected.

[0033] (3) Extraction and preprocessing of tropical cyclone precipitation fields; In this embodiment, using the intensity evolution and center latitude and longitude information of 434 rapidly intensifying tropical cyclones obtained in embodiment (2), precipitation images corresponding to the start of rapid intensification were extracted from global satellite precipitation data, with the image range set to 400 km × 400 km. Simultaneously, the environmental vertical wind shear experienced by each tropical cyclone was calculated using ERA5 reanalysis data.

[0034] Subsequently, the precipitation field was transformed from Cartesian coordinates to a polar coordinate system with the storm center as the origin through bilinear interpolation. Then, the precipitation field of each tropical cyclone was rotated according to the direction of the vertical wind shear, unifying it to a reference system where the wind shear points due north. To visually demonstrate the rotation effect, Figure 2 A schematic diagram of precipitation distribution before and after rotation is provided for a specific case. This step is performed because the precipitation structure of tropical cyclones is mainly controlled by environmental wind shear. Unifying the wind shear direction can eliminate the influence of its differences, thereby avoiding interference with subsequent clustering and classification results.

[0035] (4) Self-organizing mapping clustering classification of tropical cyclone precipitation images; In this embodiment, the precipitation images of 434 tropical cyclones in the North Pacific and North Atlantic during the period of 2000–2020 obtained in embodiment (3) were clustered and classified using a self-organizing map neural network. First, a random weight matrix with data ranging from [0,1] was generated within the program to ensure the diversity of the initial state of the neuron distribution and to avoid premature bias towards a specific pattern during training. Then, the precipitation images of the 434 tropical cyclones were standardized to eliminate the differences in overall precipitation intensity among different typhoon cases. The standardized data was input into the algorithm, and different numbers of clusters were set to obtain the corresponding classification results. Figure 3The results are presented when the total number of clusters is 2–5. It can be seen that the precipitation structure of each cluster is highly distinguishable, indicating that the self-organizing mapping method can effectively identify the precipitation structure characteristics of rapidly intensifying tropical cyclones.

[0036] Furthermore, the average variance ratio under different total number of clusters was calculated ( VR The results are shown in Table 2. When the number of clusters increases from 2 to 4, VR The value increases by approximately 4%–5% with each additional cluster; when the number of clusters increases from 4 to 5, VR The increase in value drops significantly to approximately 1%, indicating that the improvement in the difference between clusters is limited at this point. Therefore, this embodiment determines the optimal number of clusters to be 4. Figure 3 As shown in f–i, the precipitation structures of the four types of TCs exhibit the following characteristics: The first type is relatively symmetrical and compact, and therefore can be named the Symmetrical (SYM) Rapid Intensification TC; the latter three types of TCs show obvious asymmetry in their structures, and the strongest precipitation is located in the left quadrant of the reverse shear line, the left quadrant of the parallel shear line, and the right quadrant of the reverse shear line, respectively. Based on these characteristics, we name these three groups USL, DSL, and DSR type TCs, respectively. The number of rapidly intensifying TC samples in the four clusters are 112, 107, 138, and 77, respectively, accounting for 26%, 25%, 32%, and 17% of the total samples. Table 2. Total number of clusters for different clusters VR value

[0037] (5) Analysis of the convection and dynamic characteristics of different clusters; In this embodiment, based on the four classification results obtained in embodiment (4), the convective and dynamic characteristics of different clusters were analyzed using a synthetic analysis method. The statistical results of various factors for each type of tropical cyclone at the rapid intensification onset (RI onset) and the pre-intensification period (-24 h to -12 h) are as follows: Figure 4 As shown in the figure, in terms of maximum precipitation intensity, all four types of tropical cyclones experienced a significant increase in precipitation intensity before their rapid intensification, ranging from 10 to 20 degrees Celsius. The increase in convective intensity is one of the driving factors for the rapid intensification of tropical cyclones. Meanwhile, the precipitation intensity at the RI onset time of USL, DSL, and USR type TCs was 10 times higher than that of the SYM type TC. The left and right figures indicate that tropical cyclones with higher asymmetry require stronger convection to trigger rapid intensification. In terms of convective symmetry, SYM and USL type tropical cyclones underwent precipitation symmetry before rapid intensification, while DSL and DSR types did not.

[0038] From a dynamic perspective, SYM-type tropical cyclones exhibit greater intensity, smaller tilt, and smaller scale. USL-type TCs initially have a tilt of approximately 35 km, decreasing to 20 km over time; this evolution is accompanied by a process of convective symmetry. Furthermore, the intensity and scale of USL-type TCs are close to average. Compared to the previous two types of TCs, DSL and DSR-type tropical cyclones consistently maintain a large tilt (>30 km), and the DSR-type tropical cyclone is the weakest of the four types, with a median maximum wind speed of only 35 kt. These facts indicate that the rapidly intensifying tropical cyclones in the four clusters not only differ in the horizontal distribution structure of convection but also exhibit significant differences in other convective characteristics (precipitation intensity, symmetry) and dynamic characteristics (tropical cyclone intensity, scale, tilt).

[0039] (6) Environmental factor pretreatment and analysis Based on the method in step S6, this embodiment calculates multiple large-scale environmental factors, including vertical wind shear, sea surface temperature (SST), latent heat flux (LHF), mid-level relative humidity (RH), and 200 hPa upper-level divergence (D200). The scalar field rotation method in embodiment (3) is used to unify the two-dimensional fields of each environmental factor to a reference frame in which "wind shear points due north".

[0040] A preliminary comparative analysis of the regional average values ​​of various environmental factors was conducted, and the results are as follows: Figure 5 As shown. It is important to note that the vertical wind shear is an average value over a radius of 400–800 km, while the other environmental factors are average values ​​over a radius of 0–500 km. The average environmental vertical wind shear intensity for SYM category is only 5.2. ( Figure 5 The value of category a) was significantly lower than the average of other categories (95% confidence level). This result is consistent with previous research findings that tropical cyclones with greater intensity and smaller environmental vertical wind shear tend to exhibit stronger axisymmetry. The average maximum wind speed of USL-type tropical cyclones is similar to that of SYM-type cyclones, with an average environmental vertical wind shear intensity of 6. At a moderate level, the relatively strong wind shear corresponds to a more asymmetric convective structure. In contrast, the mean maximum wind speed distribution of DSL-type tropical cyclones is similar to that of USL and SYM-type cyclones, but their environmental vertical wind shear is the strongest among the four types of tropical cyclones (7.1). This corresponds to the large convective asymmetry and cyclonic tilt structure maintained before its rapid intensification. The mean environmental vertical wind shear of DSR-type tropical cyclones is 5.5. This is similar to the USL category. Therefore, the DSR category may represent a special type of tropical cyclone that enters a rapid intensification phase with relatively weak tropical storm intensity under moderate environmental wind shear conditions. Furthermore, its highly asymmetrical convective structure may also be influenced by other factors.

[0041] SYM tropical cyclones face the most unfavorable environmental conditions among the four types (except for vertical wind shear), with the lowest sea surface temperature, latent heat flux, mid-tropospheric relative humidity, and 200 hPa divergence. Figure 5 The values ​​for b–e in the figure are 28.6℃ and 165℃, respectively. 61% and Compared to the SYM category, tropical cyclones of the USL, DSL, and DSR categories have more favorable environmental conditions beyond vertical wind shear, including higher SST, LHF, mid-level RH, and D200. Specifically, the USL category is characterized by near-average sea surface temperature and mid-level relative humidity, but higher D200 and LHF; the DSL category exhibits near-average D200 and LHF, but higher sea surface temperature and mid-level relative humidity; while the DSR category has the highest sea surface temperature, D200, and mid-level relative humidity among the four categories.

[0042] The results indicate that rapidly intensifying tropical cyclones with greater intensity and weaker vertical wind shear are less dependent on other favorable environmental conditions compared to weaker tropical cyclones in environments with stronger wind shear.

[0043] (7) Correlation analysis of TC precipitation structure and environmental factors The correlation between various environmental factors and convective characteristic indicators of various tropical cyclones was assessed using the Pearson correlation coefficient, as shown in the following formula:

[0044] Figure 6The results of the correlation analysis are presented, showing that the precipitation symmetry index (PAI) is significantly positively correlated with vertical wind shear level, and significantly negatively correlated with sea surface latent heat flux and upper-level divergence, with correlation coefficients of 0.197, -0.111, and -0.114, respectively. This indicates that the stronger the vertical wind shear, the higher the sea surface latent heat flux, and the stronger the upper-level divergence, the more symmetrical the tropical cyclone will be at the moment of rapid intensification and initiation; conversely, the less symmetrical the conditions, the more asymmetrical the conditions. Furthermore, the correlation between PAI and sea surface temperature and mid-level relative humidity is not significant, with correlation coefficients of 0.001 and -0.025, respectively. Furthermore, the maximum precipitation intensity was significantly correlated with vertical wind shear, sea surface temperature, latent heat flux, and mid-level relative humidity, with correlation coefficients of 0.146, 0.219, 0.219, and 0.155, respectively, while the correlation with upper-level divergence was only 0.03. This indicates that the stronger the vertical wind shear and the higher the sea surface temperature, latent heat flux, and mid-level relative humidity, the stronger the maximum precipitation intensity of the tropical cyclone at the moment of rapid intensification.

[0045] (8) Spatial modal analysis of large-scale environmental factors Large-scale environmental factor synthesis analysis was conducted on individual cases of different types of tropical cyclones. Based on the clustering and typing results of 434 tropical cyclones in the North Pacific and North Atlantic from 2000 to 2020 in Example (4), and the preprocessed data of large-scale environmental factors in Example (6), the environmental fields corresponding to various types of rapidly intensifying tropical cyclones were synthesized, and the spatial distribution of the environmental field differences between different groups was calculated, thereby assessing the impact of environmental differences on convective structure and further revealing the causes of differences in precipitation structure of different rapidly intensifying tropical cyclones. Given that the symmetrical and compact convective structure of SYM-type TCs is mainly controlled by weak vertical wind shear and has relatively little dependence on other environmental factors, this example focuses on analyzing the distribution characteristics of environmental factors of USL, DSL, and DSR-type TCs.

[0046] Figure 7 The two-dimensional distribution characteristics of environmental factor fields at various scales for USL-type TC are shown. The mean RH of the USL-type is not significantly different from that of other categories. Figure 7 (a, d) indicates that the average mid-level RH conditions are sufficient to support precipitation symmetry. A large-scale positive LHF anomaly exists in the right quadrant of the shear direction 24–12 hours before the RI event. Figure 7 (b) These significant positive LHF anomalies are further enhanced at the rapid enhancement initiation time (exceeding 30° in the right quadrant of the cis-shear line). ), and expands in a counterclockwise direction ( Figure 7 (e in the text). Figure 7The g–j values ​​in the table show the 850 hPa wind anomaly of the USL-type TC relative to other categories within a 4000 km × 4000 km area. The most significant feature of the background anomalous circulation of the USL-type TC is the presence of a large-scale cyclonic circulation with a diameter exceeding 1500 km. Within a 500 km radius, the right quadrant of the shear direction lies within the strong cyclonic wind region on the periphery of this circulation, while the left quadrant is located in the weak wind region near the circulation center. This results in an asymmetric distribution of the total wind speed anomaly, and consequently, a significant positive LHF anomaly in the USL-type TC to the right of the shear direction. Therefore, the convective axisymmetry of the USL-type TC is mainly caused by the anomalously high sea surface heat flux driven by the large-scale cyclonic circulation surrounding the TC.

[0047] Figure 8 The two-dimensional distribution characteristics of environmental factor fields at various scales of DSL-type TC are shown. The difference field of mid-level RH during the period from -24 to -12 h ( Figure 8 a) shows a significant positive value of 2%–3% within a radius of 250–500 km to the left of the shear line. This positive value region expands over time and gradually approaches the center of the TC (Treatment Center). Figure 8 (d) For LHF, 10–20 psi were observed on the left side of the shear line during the period from -24 to -12 h. Significant positive outliers ( Figure 8 (b) At the RI initiation time, LHF showed a significant enhancement in the left quadrant of the cis-shear ( Figure 8 (e). The above results indicate that DSL-type TCs may be moving towards a wetter region with stronger latent heat flux at the sea surface before their rapid intensification, which is conducive to convection development on the left side of the shear line. In the upper troposphere, DSL-type TCs exhibit a significant positive D200 anomaly within a radius of 250–500 km in the shear line quadrant at the time of RI initiation. Figure 8 The asymmetric distribution pattern of upper-level divergence may favor convection development in the parallel-shear quadrant. Furthermore, at the onset of the RI (River Induction), a significant wind anomaly pointing towards the TC (Trunk Center) center can be observed in the lower levels along the parallel-shear direction. Figure 8 The anomalous radial inflow (i–j) may lead to enhanced inward transport of angular momentum, thereby increasing the total 10 m wind speed on the left downstream of the shear line and resulting in a stronger LHF. Furthermore, it may directly enhance boundary layer convergence, thus favoring convection development on the left downstream of the shear line.

[0048] Figure 9The two-dimensional distribution characteristics of environmental factor fields at various scales of the DSR-type TC are presented. The maximum precipitation on the right side of the shear line of the DSR-type TC may originate from the combined influence of moderate-intensity vertical wind shear and other environmental factors. In fact, the environmental conditions of the DSR-type TC are the most favorable among the four clusters, with a significant positive relative humidity anomaly (4%–5%) within a 500 km radius. Figure 9 (a, d) and upper-level divergence (values ​​are) ; Figure 9 The c and f anomalies are located in the shear quadrant, and the maximum values ​​of relative humidity and upper-level divergence are both located in the shear quadrant, which is conducive to the development of strong convection in this quadrant. Meanwhile, Figure 9 b and Figure 9 The data in section e shows that there are significant positive LHF anomalies (values ​​of 10–30) in the reverse shear quadrant and the right quadrant of the cis shear quadrant of DSR-type TC. ) and negative abnormalities (10–20) This leads to enhanced and suppressed downstream convection, respectively, causing the strongest precipitation to occur on the right side of the shear line. Similar to USL and DSL, the LHF anomaly of the DSR-type TC is also mainly controlled by the 10 m total wind speed anomaly. Figure 9 (g–j in the text). Therefore, the unique precipitation structure of the DSR-type rapidly enhanced TC is the result of the combined effects of powerful environmental dynamics, thermodynamic conditions, and specific large-scale circulation patterns.

[0049] Please see Figure 10 , Figure 10 This is a schematic diagram of the hardware device in operation according to an embodiment of the present invention. The hardware device specifically includes: a rapid enhancement type tropical cyclone precipitation structure classification and environmental impact assessment device 401, a processor 402, and a storage device 403.

[0050] A rapid enhancement tropical cyclone precipitation structure classification and environmental impact assessment device 401: The rapid enhancement tropical cyclone precipitation structure classification and environmental impact assessment device 401 realizes the rapid enhancement tropical cyclone precipitation structure classification and environmental impact assessment method.

[0051] Processor 402: The processor 402 loads and executes the instructions and data in the storage device 403 to implement the method for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment.

[0052] Storage device 403: The storage device 403 stores instructions and data; the storage device 403 is used to implement the method for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment.

[0053] The beneficial effects of this invention are: (1) This invention proposes an objective classification method for the precipitation structure of rapidly intensifying tropical cyclones, which can identify a variety of convection patterns that may occur during the rapid intensification process, and helps to deepen the understanding of the structural characteristics of rapidly intensifying tropical cyclones and the diversity of their physical mechanisms.

[0054] (2) This invention combines large-scale environmental analysis of different types of tropical cyclones to reveal the control effect of environmental factors on various precipitation structures and clarify the environmental impact mechanism of different rapid intensification types, providing new ideas and directions for improving rapid intensification forecast models and enhancing the accuracy of tropical cyclone intensity forecasts.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for rapidly enhancing tropical cyclone precipitation structure classification and environmental impact assessment, characterized in that: Includes the following steps: Step S1: Data Acquisition; Collect measured optimal track data of tropical cyclones, global satellite precipitation data, and meteorological reanalysis data; Step S2: Screening of rapidly intensifying tropical cyclones; Using the optimal track data of tropical cyclones obtained in Step S1, screen for rapidly intensifying tropical cyclones that meet the criteria based on the objective definition of rapidly intensifying events, and extract their intensity and location information. Step S3: Extraction and preprocessing of tropical cyclone satellite precipitation images; Combining the location information of rapidly intensifying tropical cyclones obtained in Step S2, precipitation images at the moment of rapid intensification of tropical cyclones are extracted from global satellite precipitation data; Subsequently, based on the meteorological reanalysis data obtained in Step S1, the environmental vertical wind shear experienced by each tropical cyclone case is calculated, and the precipitation images are rotated according to the direction of vertical wind shear to unify them into a precipitation distribution field under a reference system with the wind shear pointing due north; Step S4: Self-organizing map clustering of precipitation images; Self-organizing map neural network clustering technology is used to objectively classify the precipitation images of rapidly intensifying tropical cyclones obtained in Step S3 to clarify the possible precipitation structure types of rapidly intensifying tropical cyclones. Step S5: Analysis of convective and dynamic characteristics of individual tropical cyclones of different types; Based on the tropical cyclone clustering results of Step S4, combined with satellite precipitation data and optimal tropical cyclone track data, the convective and dynamic characteristics of tropical cyclones in different clusters are statistically analyzed, and the differences between different tropical cyclone groups are compared. Step S6: Large-scale environmental factor preprocessing; Based on the meteorological reanalysis data obtained in Step S1, directly extract or indirectly calculate the large-scale environmental variables near individual tropical cyclones, and based on the method in Step S3, rotate the large-scale environmental variable field to unify it into a variable distribution field under a reference system with the wind shear pointing due north. Step S7: Correlation analysis of the causes of differences in tropical cyclone precipitation structure; combined with the large-scale environmental variable field obtained in Step S6, calculate the mean values ​​of environmental factors in the activity area of ​​rapidly intensifying tropical cyclones, and then use the Pearson correlation coefficient to calculate the correlation between each environmental factor and the tropical cyclone convection characteristic index obtained in Step S5. Step S8: Perform large-scale environmental factor synthesis analysis on individual cases of different types of tropical cyclones; based on the clustering and typing results of step S4, synthesize the environmental fields corresponding to each type of rapidly intensifying tropical cyclone, calculate the spatial distribution of environmental field differences between different groups, and then evaluate the impact of spatial differences in environmental fields on convective structure, and clarify the causes of differences in precipitation structure of different rapidly intensifying tropical cyclones.

2. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: In step S1, the optimal path data of the tropical cyclone includes the center location of the tropical cyclone and the maximum near-sea surface wind speed, with a data time resolution of 6 hours. The global satellite precipitation data mentioned above is high spatiotemporal resolution gridded precipitation data, with a temporal resolution of 30 minutes per hour and a spatial resolution of 0.1° × 0.1°. The meteorological reanalysis data includes meridional wind, zonal wind, atmospheric temperature, relative humidity, sea surface temperature, wind speed at 10 m altitude, and specific humidity at 2 m altitude.

3. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: In step S2, the screening criteria for rapidly intensifying tropical cyclones include three steps: S21: The rate of change of intensity of a tropical cyclone during its rapid intensification phase shall not be less than 30 knots / day; S22: The intensity of the tropical cyclone reaches the level of a severe tropical storm or above at the moment of its rapid intensification. S23: The tropical cyclone was located south of 30°N during its rapid intensification phase and did not make landfall.

4. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: In step S3, the precipitation image at the moment of rapid intensification of the tropical cyclone is extracted from global satellite precipitation data, with a range of 400 km × 400 km and the center of the tropical cyclone as the image center; the environmental vertical wind shear is defined as the difference between the 200 hPa and 800 hPa mean wind vectors within a circular range with a radius of 400 km–800 km from the typhoon center, calculated using the following formula: In the formula, VWS The magnitude of vertical wind shear. D The vertical wind shear direction is expressed in radians. , This represents the average zonal and meridional wind vectors within a radius of 400–800 km from the center of the TC. The subscripts 200 and 850 indicate that the variable is located at a level of 200 or 850 hPa. Bilinear interpolation is used to transform the precipitation field from Cartesian coordinates to a polar coordinate system with the storm center as the origin; the west direction is defined as 0 radians and the north direction as π / 2; then, the precipitation field is rotated accordingly based on the angle between the vertical wind shear direction and the north direction in each case, so as to unify the precipitation distribution field under the reference system with the wind shear pointing to the north.

5. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: Step S4, which uses a self-organizing map neural network to cluster tropical precipitation images, includes the following three steps: S41: Generate a random weight matrix and initialize it in the [0,1] interval or within the data range to ensure that the initial state of the neuron distribution is diverse and to avoid premature bias towards a specific pattern during training. S42: The precipitation rate in the precipitation image is normalized using the following formula: in X and X′ Distinguish between the original precipitation rate and the standardized precipitation rate. n This represents the total number of grid points in the two-dimensional field of precipitation rate. The purpose of this homogenization process is to eliminate the differences in overall precipitation intensity among different typhoon cases, so that clustering focuses more on the spatial distribution structure of precipitation. S43: By calculating the average variance ratio VR The optimal number of clusters is determined by the following formula: in This represents the squared distance between the centroid of a given cluster and the centroid of the total sample. This represents the mean of the squared distances between all samples in the cluster and the total number of samples from the centroid. VR A higher mean indicates more significant differences between clusters; when VR When the number of clusters reaches a certain critical number and then plateaus, that number is taken as the optimal number of clusters.

6. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: In step S5, based on the clustering results, a composite analysis is performed on the convective characteristics, including the maximum precipitation intensity and convective symmetry of tropical cyclones in various clusters, as well as the dynamic characteristics, including cyclone intensity, scale and tilt, i.e. the sample mean is calculated. Maximum precipitation intensity is defined as the maximum precipitation rate within a 200 km radius. Convective symmetry is measured using the Precipitation Asymmetry Index (PAI), calculated as follows: in R n = R 1–6 / R 0 represents the standardized precipitation rate. R 1–6 and R 0 represents the sum of the asymmetric components and the symmetric component of the precipitation rate for waves 1–6, respectively. r and λ These represent the radius and azimuth, respectively. A larger PAI value indicates a more asymmetrical precipitation structure; The intensity of a tropical cyclone is measured by the maximum wind speed near the sea surface, its scale is measured by the radius of the gust circle, i.e. the radius of the 34-knot wind circle, and its inclination is measured by the relative distance between the vortex centers at 500 hPa and 850 hPa.

7. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: Step S6, the large-scale environmental factor preprocessing, includes the following two steps: S61: Calculate the average relative humidity of the 700–500 hPa pressure layer as an assessment standard for mid-tropospheric humidity; calculate upper-level divergence based on the meridional and zonal wind fields of the 200 hPa pressure layer; calculate the sea surface latent heat flux (LHF) based on the overall aerodynamic formula, as follows: in L v Represents the latent heat of water vapor phase change. C k Represents the latent heat transfer coefficient, with a value of , ρ d The density of dry air is [value missing]. , U 10 At a height of 10m, q *and q 2 represents the saturated specific humidity at the sea surface and at a height of 2 m, respectively; S62: Extract the environmental variable field of 1000 km × 1000 km centered on the tropical cyclone, and unify it to a reference frame with the wind shear pointing due north, following the method in step S3.

8. The method for classifying the precipitation structure and assessing the environmental impact of rapidly intensifying tropical cyclones as described in claim 1, characterized in that: In step S8, when assessing the large-scale environmental conditions of different clusters of tropical cyclones, the horizontal space is divided into four quadrants relative to the direction of environmental wind shear: left side of the shear, right side of the shear, left side of the reverse shear, and right side of the reverse shear, where "forward" and "reverse" represent the direction consistent with and opposite to the direction of wind shear, respectively.

9. A storage device, characterized in that: The storage device stores instructions and data to implement the rapid enhancement method for tropical cyclone precipitation structure classification and environmental impact assessment as described in any one of claims 1 to 8.

10. A rapid-enhancing tropical cyclone precipitation structure classification and environmental impact assessment device, characterized in that: include: A processor and a storage device; the processor loads and executes instructions and data in the storage device to implement the method for rapid enhancement of tropical cyclone precipitation structure classification and environmental impact assessment as described in any one of claims 1 to 8.

Images