A Similar Typhoon Identification Method Based on Vector-Raster Analysis with Multi-parameter Constraints

Abstract

The present invention discloses a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints. That is, for a target typhoon, an optimized discrimination index for similar typhoons is established through three parameters: the typhoon center position, typhoon intensity, and typhoon moving speed. By comprehensively using the analysis methods of vector data and raster data, typhoons with similar path positions and high matching degrees of typhoon intensity and moving speed at the position closest to the typhoon center are identified. Constraint parameter combinations can be selectively applied according to specific application requirements, and similar typhoons are identified and sorted through the proposed typhoon similarity discrimination index. The present invention establishes a fast identification algorithm for similar typhoons based on three-parameter constraints of typhoon path, intensity, and moving speed. Since the information of typhoon intensity and / or moving speed can be matched simultaneously, the potential disasters of the identified typhoons to the coast are more similar, and it can be better applied to the prediction of the disaster levels such as storm surge water level rise, typhoon wave impact area, and flooding caused by similar typhoons.

Description

Technical Field

[0001] The present invention belongs to the field of earth sciences, and particularly relates to a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints. Background Art

[0002] The accurate identification of typhoons with similar paths is crucial for typhoon path prediction and disaster forecasts such as coastal storm surges, typhoon waves, and flooding. Typhoons with similar paths play an important role in the early warning of affected areas and scopes, and can effectively guide disaster prevention strategies and reduce losses of people's lives and property. However, if typhoons with similar paths do not match in terms of intensity, etc., the damage degree of corresponding disasters, such as the height of storm surge water level rise, flooding depth and scope, cannot be effectively estimated, which greatly reduces the effect of disaster damage prediction based on historical typhoon databases. The purpose of this study is to propose a calculation index and method for typhoon path similarity that simultaneously considers typhoon intensity and typhoon moving speed constraints, which can ensure that the identified typhoon paths have good similarity in terms of path position, typhoon intensity, and typhoon moving speed.

[0003] The similarity of typhoon paths can have specific connotations in different aspects. The specific parameters can include typhoon center position, intensity, moving speed, and the time when the typhoon passes through, etc. However, if all aspects are required to match, it may affect the identification efficiency and easily exclude typhoons with better matching in many features but lacking in matching in only one aspect; therefore, in practical applications, it is more crucial to balance the similarity of typhoon paths and the usability of corresponding evaluation parameters and indexes. Especially in the application of coastal disaster prevention and warning, endowing specific connotations to the similarity of typhoon paths directly affects the reasonable estimation of disasters.

[0004] Therefore, a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints is needed to solve the above problems. Summary of the Invention

[0005] Object of the Invention: In view of the defect that when evaluating coastal typhoon storm surges, typhoon waves, and flooding disasters, it is impossible to reasonably predict the disaster impact degree on the basis of considering both the disaster area and scope only by the similarity of typhoon path positions, the present invention provides a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints, and optimizes the existing typhoon similarity identification indexes and algorithms.

[0006] Technical Solution: To solve the above technical problems, a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints disclosed by the present invention adopts the following technical solution:

[0007] The present invention selectively constrains the typhoon path similarity based on three parameters: the typhoon center position, typhoon intensity, and moving speed, and proposes an effective similarity typhoon identification index including the above multi-parameter constraints and a calculation method based on vector-raster comprehensive analysis, comprising the following steps:

[0008] Step 1, establish a historical typhoon database;

[0009] Step 2, preliminarily identify historical typhoons with typhoon path positions similar to the target typhoon, and form a subset of typhoons to be identified;

[0010] Step 3, for each typhoon to be identified in the subset of typhoons to be identified, calculate the nearest position and the nearest distance corresponding to each typhoon center position on the target typhoon path, and form a pairing relationship for all typhoon center positions of the target typhoon;

[0011] Step 4, for each typhoon to be identified in the subset of typhoons to be identified, calculate the similarity index value under the simultaneous constraints of multiple parameters such as typhoon path position, intensity, and / or moving speed with the target typhoon, obtain historical typhoons with similar combined constraint parameters, and sort them.

[0012] Furthermore, Step 1 includes: establishing a historical typhoon database based on the typhoon best track dataset, which includes the longitude and latitude of the typhoon center position, typhoon level, and maximum typhoon wind speed. The historical typhoon database also includes the typhoon path length and typhoon moving speed data obtained through calculation. Among them, the typhoon best track dataset uses data publicly obtained from meteorological agencies.

[0013] Furthermore, Step 2 includes:

[0014] Step 2.1, take the target typhoon path as a "line" vector object, perform buffer calculation with L as the width, and obtain the buffer polygon P;

[0015] Step 2.2, traverse all typhoons to be identified in the historical typhoon database, and count the percentage of typhoon centers of each typhoon to be identified located in the polygon P

[0016] where N in is the number of typhoon centers of a certain typhoon to be identified located inside the polygon P, and N total is the total number of typhoon centers of the typhoon to be identified;

[0017] Step 2.3, when the path length ratio between the target typhoon and the typhoon to be identified satisfies ratio len ≥R1, ratio len ≤1 / R1, and ratio loc ≥R2, it is preliminarily identified as having a similar typhoon path position and included in the subset of typhoons to be identified;

[0018] Among them, R1 represents the similarity threshold of typhoon path length, and R2 represents the percentage threshold of the typhoon center position on the typhoon path to fall into the polygon P.

[0019] In the buffer polygon P, jointly constrained by the typhoon path length similarity threshold R1 and the percentage threshold R2 of the typhoon center position on the typhoon path to fall into the polygon P, traverse the typhoon path database to identify typhoons with similar path positions, and preliminarily screen the typhoon to be identified through the path position, which can effectively reduce the number of typhoons in the subset. Subsequently, the traversal calculation for typhoons does not need to be performed on all historical typhoons, which can improve the efficiency of identifying similar typhoons.

[0020] Furthermore, step 4 includes calculating the similarity of the intensities of the typhoon to be identified and the target typhoon. The typhoon intensity uses the maximum wind speed at different typhoon center positions as an indication of typhoon intensity, and the unit of the maximum wind speed is m / s; calculating the similarity of the intensities of the typhoon to be identified and the target typhoon includes the following steps:

[0021] Step 4.1, interpolate the typhoon center position and the maximum wind speed of the typhoon to be identified with Δt to form typhoon path data with a time resolution of Δt. and respectively represent the spherical coordinates of the typhoon center position j at any moment after interpolation and the corresponding maximum wind speed, where j ≥ 1.

[0022] Step 4.2, based on the typhoons preliminarily identified as having similar paths, denote the typhoon center positions on the target typhoon path as and their corresponding maximum wind speeds as i represents the index of each typhoon center position, 1 ≤ i ≤ I, and I represents the total number of typhoon center positions on the target typhoon path; the closest position on the typhoon path of the typhoon to be identified obtained by calculating in step 3 is denoted as J i , then the difference in typhoon intensity between the paired typhoon center positions is The index reflecting the intensity similarity between the two typhoon paths is defined as The subscripts o and s in the above formula correspond to the target typhoon path data and the typhoon path data to be identified after interpolation, respectively.

[0023] The pairing method of the typhoon center position proposed by the present invention completely overcomes the limitations and unreasonable aspects of the existing algorithms, and is also the key to the evaluation of the similarity of typhoon path positions, intensity similarities, and traveling speed similarities; in addition, the intensity similarity index calculation formula based on the maximum wind speed difference has the advantages of strong representativeness and simple calculation. The relevant beneficial effects will be further described in specific embodiments in conjunction with the accompanying drawings.

[0024] Further, step 4 also includes calculating the similarity between the typhoon to be identified and the target typhoon in terms of their moving speeds, which includes the following steps:

[0025] Step 4.3, interpolate the moving speed of the typhoon to be identified at intervals of Δt, representing the moving speed of the typhoon center position j at any moment after interpolation;

[0026] Step 4.4, based on the typhoons preliminarily identified as having similar paths, denote the moving speeds of the typhoon center positions on the target typhoon path as Then the difference in the moving speeds of the typhoons at the paired typhoon center positions is The index reflecting the similarity of the moving speeds of the two typhoons on their paths is defined as

[0027] The similarity indices of the typhoon path position, intensity, and moving speed proposed by the present invention are consistent in form, and the cumulative number of each index depends on the number of typhoon center positions of the target typhoon, which is also exactly the same. By using the similarity indices of each parameter proposed by the present invention, they can balance each other in magnitude to ensure that the effects are effectively exerted in identifying typhoon similarity.

[0028] Further, the calculation of the typhoon similarity index based on multi-parameter constraints proposed in step 4 is based on the determination of the closest positions of the typhoon centers, forming a one-to-one pairing relationship between the target typhoon and the typhoon to be identified in terms of their center positions, and calculating the matching indices for the typhoon path position, intensity, and moving speed; the specific formula for the comprehensive typhoon similarity index is:

[0029]

[0030] where d i is the closest distance from any typhoon center position of the target typhoon to the path of the typhoon to be identified, with the unit of Km; D start represents the spherical distance between the starting point of the path of the typhoon to be identified and the target typhoon path, and D end represents the spherical distance between the ending point of the path of the typhoon to be identified and the target typhoon path, with the unit of Km; () represents whether the parameter constraint effect can be selectively applied or not. The contents in the brackets in the formula are the typhoon intensity constraint and the moving speed constraint in sequence. A is the weight coefficient of the moving speed term, which can be adjusted appropriately. The remaining three terms are all constraints on the typhoon path position; to ensure a better typhoon similarity identification effect, the calculation results of the relevant parameters such as typhoon position, typhoon intensity, and moving speed in this index need to use the units marked above.

[0031] The present invention proposes a similarity recognition method that selectively combines three parameters: the typhoon path position, intensity, and moving speed. This method not only covers the three most critical parameters affecting typhoon storm surge coastal disaster forecasting and assessment but also avoids the problem that considering too many parameters may instead affect the recognition effect and efficiency of similar typhoons. The relevant beneficial effects will be further described in specific embodiments. At the same time, the selective combination of multi-parameter similarity indicators proposed by the present invention has good flexibility. Users can select corresponding similarity parameters according to the specific needs of coastal disaster prediction or disaster prevention management. For example, when estimating the disaster risk area, the similarity of the typhoon path position is the most important. If the disaster hazard needs to be considered simultaneously, the similarity of the typhoon intensity and moving speed should also be taken into account.

[0032] The smaller the value of this indicator, the higher the degree of simultaneous matching of the typhoon path in terms of position, intensity, and / or moving speed. When the indicator is 0, the two typhoon paths are exactly the same. By sorting the values of the recognition indicator from small to large, typhoons with similar typhoon positions, intensities, and / or moving speeds from high to low can be identified.

[0033] Further, calculating the nearest position and the nearest distance corresponding to each typhoon center position on the target typhoon path in step 3 includes: interpolating and densifying the typhoon path to be recognized. For each typhoon center position on the target typhoon path, traverse each interpolated position on the typhoon to be recognized, and obtain the position J corresponding to the minimum spherical distance. i This is the nearest position of the typhoon center, and the corresponding distance is the nearest distance.

[0034] The recognition of the nearest position proposed by the present invention is based on the appropriate densification of the typhoon to be recognized. The similarity indicator calculated thereby can ensure that the position, intensity, and moving speed of the typhoon to be recognized are all similar to the target typhoon center position at the nearest position.

[0035] Further, calculating the nearest distance corresponding to each typhoon center position on the target typhoon path in step 3 includes: interpolating and densifying the typhoon path to be recognized, and covering the target typhoon path with a rectangular area; for the interpolated and densified typhoon path to be recognized, calculate the rasterized proximity within the rectangular area covered by the target typhoon path, and then the nearest distance at each center position of the target typhoon can be directly interpolated from the proximity field.

[0036] The present invention provides another algorithm for calculating the nearest distance by proposing a calculation method based on proximity, providing multiple implementation paths for users to implement the specific application of the present invention.

[0037] Further, the formation of the pairing relationship for all typhoon center positions in step 3 is achieved based on identifying the nearest positions, including finding the nearest points on the interpolated and encrypted typhoon paths to be identified for each center position on the target typhoon path, ensuring that the number of typhoon center pairing groups depends on the target typhoon, thereby achieving consistency in the calculation of indicators for each typhoon to be identified.

[0038] Existing algorithms pair the typhoon path center positions one by one and cannot well solve the problems of incomplete recording of typhoon center positions caused by differences in typhoon path lengths and the cumulative dislocation of typhoon center positions caused by differences in typhoon moving speeds. In contrast, the present invention can effectively avoid the above problems. The beneficial effects of the present invention and its differences from existing algorithms will be further described in specific embodiments in conjunction with the accompanying drawings.

[0039] Further, the typhoon path length in step 1 and the spherical distance calculations in steps 3 and 4 both adopt the great circle distance calculation.

[0040] There are no specific limitations on the specific values of the parameters such as Δt, buffer width L, typhoon path length similarity threshold R1, and percentage threshold R2 for the typhoon path data interpolation and encryption used in the present invention. These values can be adjusted according to differences in the research area or research problems.

[0041] Beneficial effects: The present invention has a clear concept and an algorithm that is easy to implement. The proposed typhoon similarity discrimination index can quickly and effectively identify historical typhoon paths that are similar in terms of typhoon path travel positions, corresponding intensities, and travel speeds. It can provide similar typhoon samples for disaster prediction agencies, effectively predict the disaster range and degree that future typhoons may cause to coastal areas, and assist local governments in making reasonable decisions on disaster prevention and avoidance.

[0042] In addition, the typhoon similarity discrimination index proposed by the present invention is also superior to existing algorithms in the typhoon path pairing algorithm. Existing algorithms pair the typhoon path center positions one by one and cannot well solve the problems of incomplete recording of typhoon center positions caused by differences in typhoon path lengths and the cumulative dislocation of typhoon center positions caused by differences in typhoon moving speeds. The discrimination index and its calculation method proposed by the present invention can well solve this problem, and the specific details will be further elaborated in the embodiments in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The following further specifically describes the present invention in conjunction with the accompanying drawings and specific embodiments. The above and / or other advantages of the present invention will become clearer.

[0044] Figure 1Schematic diagram of a multi-parameter constrained similar typhoon identification method based on vector-raster analysis provided by an embodiment of the present application.

[0045] Figure 2 Schematic diagram of calculating the nearest neighbor distance based on proximity grid data in a multi-parameter constrained similar typhoon identification method based on vector-raster analysis provided by an embodiment of the present application.

[0046] Figure 3 Schematic diagram for comparing typhoon center position pairing algorithms. a) is the existing algorithm, and b) is the algorithm of the present invention.

[0047] Figure 4 Schematic diagram of the most similar typhoon identified for Typhoon Lekima in 2019. The most similar typhoon identified is Maysak in 2015, and the legend shows the typhoon levels at different typhoon center positions.

[0048] Figure 5 Schematic diagram of the top ten similar typhoons identified for Typhoon Lekima in 2019.

[0049] Figure 6 Schematic diagram of the top ten similar typhoons identified for Typhoon Virginia in 1978 without constraining the landing characteristics.

[0050] Figure 7 Schematic diagram of the top ten similar typhoons identified for Typhoon Virginia in 1978 when constraining the landing characteristics.

[0051] Marking description: 1 - Target typhoon path, with the typhoon center positions at each moment represented by circles; 2 - Encrypted typhoon path to be identified, with the typhoon center positions at each moment represented by black dots; 3 - Nearest neighbor distance d from a certain center position of the target typhoon to the typhoon to be identified; i ; 4 - Maximum wind speed or traveling speed at the nearest position on the typhoon path to be identified that is paired with a certain center position of the target typhoon or traveling speed 5 - Maximum wind speed or traveling speed at a certain center position of the target typhoon path; 6 - Buffer width L; 7 - Buffer polygon P of the target typhoon; 9 - Proximity matrix; 10 - Contour lines of the proximity field (unit: Km); 11 - Connection line between the target typhoon center position and its paired position. Detailed implementation manner

[0052] The following further clarifies the implementation scheme of a multi-parameter constrained similar typhoon identification method based on vector-raster analysis proposed by the present invention in combination with specific embodiments.

[0053] It is a common consensus in the academic community that the typhoon path position (the typhoon landing location is included in this category) and typhoon intensity have a greater impact on typhoon storm surge disasters (such as Sebastian et al., 2014). And the typhoon moving speed and typhoon radius have also been proven to be important factors affecting coastal storm surge disasters (Rego & Li, 2009; Yang et al., 2019). Therefore, typhoon similarity should include the above connotations. In this invention, the maximum typhoon wind speed is used to characterize typhoon intensity. Both the maximum typhoon wind speed and the central minimum pressure are indicators used to characterize typhoon intensity. According to existing research (such as Yang et al., 2019), the wind speed in the offshore area plays a key role in storm surge water level rise, typhoon wave height, etc. And internationally, the maximum wind speed is generally used as the standard for typhoon classification. Using the maximum wind speed to characterize typhoon intensity is more representative. At the same time, it is noted that most of the publicly available typhoon best track datasets do not provide information on the maximum wind speed radius of typhoons. And the estimation of the typhoon radius generally uses empirical formulas, mostly establishing relationships with the maximum wind speed and central pressure difference of typhoons (such as Knaff et al., 2007; Willoughby & Rahn, 2004). Due to the uncertainty of empirical formulas, there are significant differences in the results of each formula. Therefore, if the typhoon radius information obtained by empirical formulas is incorporated into the similar typhoon identification index, it will introduce uncertainty and additional errors, affecting the identification effect of similar typhoons. The typhoon moving speed can be calculated based on the typhoon center position in the typhoon best track dataset. Additionally, it should be pointed out that parameters such as the typhoon occurrence time (such as month), generation source location, etc. are not important for evaluating and predicting the scope and degree of typhoon offshore storm surge disasters. Therefore, they are not included in the typhoon similarity index of this invention.

[0054] References: Knaff, J., Sampson, C., DeMaria, M., Marchok, T., Gross, J., & McAdie, C. (2007). Statistical tropical cyclone wind radii prediction using climatology and persistence. Weather and Forecasting, 22(4), 781–791. https: / / doi.org / 10.1175 / WAF1026.1

[0055] Rego, J.L., & Li, C. (2009). On the importance of the forward speed of hurricanes in storm surge forecasting: A numerical study. Geophysical Research Letters, 36(7), L07609. https: / / doi.org / 10.1029 / 2008GL036953

[0056] Sebastian, A., Proft, J., Dietrich, J.C., Du, W., Bedient, P.B., & Dawson, C.N. (2014). Characterizing hurricane storm surge behavior in Galveston Bay using the SWAN+ADCIRC model. Coastal Engineering, 88(Journal Article), 171–181. https: / / doi.org / 10.1016 / j.coastaleng.2014.03.002

[0057] Willoughby, H.E., & Rahn, M.E. (2004). Parametric Representation of the Primary Hurricane Vortex. Part I: Observations and Evaluation of the Holland (1980) Model. Monthly Weather Review, 132(12), 3033–3048. https: / / doi.org / 10.1175 / MWR2831.1

[0058] Yang, J., Li, L., Zhao, K., Wang, P., Wang, D., Sou, I. M., et al. (2019). A Comparative Study of Typhoon Hato (2017) and Typhoon Mangkhut (2018) — Their Impacts on Coastal Inundation in Macau. Journal of Geophysical Research: Oceans, 124(12), 9590–9619. https: / / doi.org / 10.1029 / 2019JC015249

[0059] In summary, a similarity typhoon identification method based on vector-raster analysis with multi-parameter constraints proposed in this embodiment considers three key parameters: typhoon path position, typhoon intensity, and traveling speed. It not only covers important influencing factors for evaluating coastal storm surge disasters but also achieves the purpose of rapid identification. At the same time, by optimizing the similarity discrimination index, the existing discrimination algorithm is optimized. The specific steps are as follows:

[0060] Step 1: Establish a historical typhoon database, including:

[0061] The typhoon best track dataset can use any publicly available data. The typhoon best track dataset in this embodiment is formed based on the typhoon best track data of the China Meteorological Administration (CMA). This dataset contains typhoon path data since 1949, with data recorded every 6 hours. Each group of records specifically includes the date and time when the typhoon center occurred, the longitude and latitude of the typhoon center position, the typhoon level, the maximum typhoon wind speed, and the central minimum air pressure.

[0062] On this basis, calculate the typhoon path length (unit: Km) and the typhoon traveling speed (unit: Km / h) and form a historical typhoon database together with the above data.

[0063] Step 2: Initially identify historical typhoons with typhoon path positions similar to the target typhoon and form a subset of typhoons to be identified, specifically including:

[0064] Step 2.1: Take the target typhoon path as a "line" vector object and perform buffer calculation with L as the width to obtain a buffer polygon P.

[0065] Step 2.2: Traverse all typhoons to be identified in the historical typhoon database and count the percentage ratio of the typhoon centers of each typhoon to be identified located in the polygon P. loc ;

[0066] Step 2.3, when the path length ratio between the target typhoon and the typhoon to be identified satisfies ratio len ≥R1, ratio len ≤1 / R1, and ratio loc ≥R2, it is preliminarily identified that the typhoon path positions are similar and included in the subset of typhoons to be identified;

[0067] Among them, R1 represents the typhoon path length similarity threshold, and R2 represents the percentage threshold of the typhoon center position on the path of the typhoon to be identified falling into the polygon P.

[0068] Constrained by the typhoon path length similarity threshold R1 and the percentage threshold R2 of the typhoon center position on the path of the typhoon to be identified falling into the polygon P, traverse the historical typhoons in the historical typhoon database, identify the historical typhoons with similar path positions, and form a subset of typhoons with similar paths, as Figure 1 shown. The calculation formulas of the above two discrimination indexes and their thresholds are as follows:

[0069]

[0070]

[0071] In this embodiment, the value of R1 is 0.6, and the value of R2 is 0.7, that is

[0072] ratio len ≥0.6 and ratio len ≤1.6667 and ratio loc ≥0.7 (3)

[0073] Among them, Len is the typhoon path length, and the subscripts o and s correspond to the target typhoon path data and the path data of the sample typhoon to be identified respectively, then Len o and Len s are the path lengths of the target typhoon and the typhoon to be identified respectively; N total and N in are the total number of records of the typhoon center position to be identified and the number of records falling into the polygon P respectively.

[0074] Step 3, determination of the closest distance and position of the typhoon center

[0075] In order to more accurately discriminate the closest points between two typhoons, interpolate the path position, maximum wind speed and traveling speed of the typhoon to be identified with Δt = 1h to form typhoon path data with a time resolution of Δt. The interpolation can be based on linear interpolation.

[0076] In an alternative implementation, as Figure 1 shown, for each center position on the target typhoon path Traverse each interpolation position on the typhoon to be recognized, and obtain the position J corresponding to the minimum spherical distance i , and the corresponding distance is the nearest distance d i (Unit: Km).

[0077] In another alternative implementation, the determination of the nearest distance of the typhoon center can also be carried out in the following way. Specifically: for the path of the typhoon to be recognized after interpolation encryption, calculate the rasterized proximity (unit: Km) within the rectangular area covered by the target typhoon path. Then, the nearest distance at each center position of the target typhoon can be directly interpolated from the proximity field, as Figure 2 shown.

[0078] Either of the two methods for determining the nearest distance of the typhoon center proposed in this embodiment can be selected. When the spacing of the rasterized proximity is reasonably selected, the speeds of the two algorithms are comparable. On a personal computer with a CPU of 32G and a Windows 7 (64-bit) system, the total running time including operation and drawing of relevant pictures does not exceed 1 minute; the results of the two algorithms for recognizing similar typhoons are exactly the same.

[0079] Step 4, Similar typhoon recognition and sorting based on multi-parameter constraints

[0080] In the present invention, the center positions on the path of the target typhoon are paired one by one through the nearest positions of the typhoon center, and the matching indexes of the typhoon path position, intensity and traveling speed are calculated. The difference in the pairing effect from the existing algorithm can be seen in Figure 3 .

[0081] Based on the subset of typhoons to be recognized with similar paths preliminarily recognized in Step 2, perform loop calculations for each typhoon. For each typhoon center position on the target typhoon path and its corresponding maximum wind speed and traveling speed Determine the nearest position number as J by Step 3 i , then the differences in typhoon intensity and traveling speed between the paired typhoon center positions are respectively and From this, the typhoon path similarity index S of this typhoon can be further calculated. The calculation formula for the typhoon similarity index considering multiple parameters of typhoon path position, intensity and traveling speed is as follows:[[]]

[0082]

[0083] where I is the number of records of typhoon center positions in the target typhoon path data; d i is the nearest distance between any typhoon center position of the target typhoon and the path of the typhoon to be recognized, unit Km; The maximum typhoon wind speed at the nearest location paired with the target typhoon center position, unit: m / s; The typhoon moving speed at the nearest location paired with the target typhoon center position, unit: Km / h; D start The spherical distance between the typhoon path to be identified and the starting point of the target typhoon path, unit: Km; D end The spherical distance between the typhoon path to be identified and the ending point of the target typhoon path, unit: Km; () represents whether the parameter constraint effect can be selectively applied or not. In the formula, the contents in the brackets are the typhoon intensity constraint and the moving speed constraint in sequence, and the remaining three items are all constraints on the typhoon path position; A is the weight coefficient of the moving speed term, which can be appropriately adjusted, and is taken as 1 in this embodiment.

[0084] Sorting in ascending order according to the value of the similarity index S can identify typhoons with comprehensive similarity from high to low in terms of typhoon position, intensity and moving speed; when this index is 0, the two typhoon paths are completely the same. Sorting in ascending order according to the value of the identification index can identify typhoons with similarity from high to low. For example, in this embodiment, when the 2019 severe typhoon Lekima is used as the target typhoon, the most similar typhoon identified is the 2015 typhoon Matsa, as Figure 4 shown, the typhoon paths of the top ten in similarity ranking are as Figure 5 shown.

[0085] In the similarity typhoon identification method proposed by the present invention, other constraints related to the typhoon path position can be further imposed. For example, if it is desired that the selected similar typhoons have the same coastal landing characteristics as the target typhoon, the discrimination of whether the typhoon lands or not can be added on the basis of the above algorithm. Figure 6 and Figure 7 show the top ten similar path typhoons identified for the 1978 typhoon Virginia (this typhoon did not land in Japan) without and with constraints on the typhoon landing characteristics respectively. Because additional landing constraint conditions are imposed, there are certain differences in the identified typhoons.

[0086] Still taking the 2019 typhoon Lekima as an example, according to a multi-parameter constraint-based similar typhoon identification method based on vector-raster analysis proposed by the present invention, the top 10 identified typhoons are shown in the following table. Since the typhoon Lekima itself is also in the historical typhoon library, no matter how the combination of selectively imposed constraint parameters is carried out, the first similar typhoon identified is itself, see the typhoon numbered 0. The calculation results also show that the identification results obtained by the two calculation methods of the nearest distance between typhoon centers proposed in step 3 are completely the same; whether the typhoon moving speed is imposed has an impact on the identified typhoons and their ranking.

[0087] Specifically, regarding the selective combination of multi-parameter similarity indicators proposed in the present invention, users can select corresponding similarity parameters according to the specific requirements of coastal disaster prediction or disaster prevention management. For example, when estimating the disaster risk area, the similarity of the typhoon path position is the most important. If the similarity of typhoon intensity and moving speed is to be considered simultaneously when considering disaster hazard, it should also be taken into account. Generally speaking, between typhoon intensity and moving speed, the former plays a more crucial role because the moving speed of a typhoon is usually within the range of 15 - 30 Km / h, which is much smaller than the variation range of the maximum wind speed of the typhoon. Therefore, the disaster differences caused by the moving speed between different typhoons are usually smaller than those caused by typhoon intensity. So when the moving speed is relatively important, its similarity must be considered. Otherwise, the similarity discrimination only including typhoon path position and intensity generally can also meet the application requirements.

[0088] Table 1 Comparison of the sorted similar typhoons identified for Typhoon Lekima in 2019

[0089]

[0090] Note: Algorithm 1 refers to calculating the nearest neighbor distance of the typhoon center by traversing and interpolating to encrypt the path position; Algorithm 2 refers to calculating the nearest neighbor distance by proximity.

[0091] In specific implementation, the present application provides a computer storage medium and a corresponding data processing unit. Among them, the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, it can run the content of the invention of a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints and some or all of the steps in each embodiment. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0092] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of a computer program and its corresponding general hardware platform. Based on such an understanding, the essence of the technical solutions in the embodiments of the present invention, or the part that contributes to the prior art, can be embodied in the form of a computer program, that is, a software product. The computer program software product can be stored in a storage medium, including several instructions to enable a device (which can be a personal computer, a server, a single-chip microcomputer, a MUU, or a network device, etc.) containing a data processing unit to execute the methods described in each embodiment or some parts of the embodiments of the present invention.

[0093] The present invention provides a method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints. There are many methods and approaches to specifically implement this technical solution. The above description is only a specific implementation manner of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. Each component not clearly defined in this embodiment can be implemented by using existing technologies.

Claims

1. A method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints, characterized in that It includes the following steps: Step 1, establish a historical typhoon database; Step 2, preliminarily identify historical typhoons with typhoon path positions similar to that of the target typhoon, and form a subset of typhoons to be identified; Step 3, for each typhoon to be identified in the subset of typhoons to be identified, calculate the nearest position and the nearest distance corresponding to each typhoon center position on the target typhoon path, and form a pairing relationship for all typhoon center positions of the target typhoon; Step 4, for each typhoon to be identified in the subset of typhoons to be identified, calculate the similarity index value under the simultaneous constraints of multiple parameters such as typhoon path position, intensity, and / or moving speed with respect to the target typhoon, obtain historical typhoons with similar combination constraint parameters, and sort them; Step 4 includes calculating the similarity of the intensity between the typhoon to be identified and the target typhoon. The typhoon intensity uses the maximum wind speed at different typhoon center positions as an indication of typhoon intensity, and the unit of the maximum wind speed is m / s. The calculation of the similarity of the intensity between the typhoon to be identified and the target typhoon includes the following steps: Step 4.1, interpolate the typhoon center position and maximum wind speed of the typhoon to be recognized at intervals of Δt to form typhoon path data with a time resolution of Δt. and w sj respectively represent the spherical coordinates of the typhoon center position j at any moment after interpolation and the corresponding maximum wind speed, where j ≥ 1. Step 4.2, based on the typhoons initially identified as having similar paths, record the positions of the typhoon centers on the target typhoon path as and their corresponding maximum wind speeds as where i represents the index of each typhoon center position, 1 ≤ i ≤ I, and I represents the total number of typhoon center positions on the target typhoon path; the nearest position on the typhoon path to be identified obtained in Step 3 is denoted as J i , then the typhoon intensity difference between the paired typhoon center positions is The index reflecting the intensity similarity between the two typhoon paths is defined as In the above formula, the subscripts o and s correspond to the interpolated target typhoon path data and the typhoon path data to be identified, respectively; Step 4 also includes calculating the similarity of the moving speed between the typhoon to be identified and the target typhoon, which includes the following steps: Step 4.3, interpolate the moving speed of the typhoon to be recognized with Δt as well, which represents the moving speed of the typhoon center position j at any moment after interpolation; Step 4.4, based on the typhoons initially identified as having similar paths, record the traveling speeds of the typhoon centers at each position on the target typhoon path as Then, the difference in the traveling speeds of the typhoons at the paired typhoon center positions is The index reflecting the similarity of the traveling speeds of the two typhoons on their paths is defined as 2. The similarity typhoon recognition method based on vector-raster analysis with multi-parameter constraints according to claim 1, wherein, Step 1 includes: based on the typhoon best track dataset, establish a historical typhoon database including typhoon center longitude and latitude, typhoon level, typhoon maximum wind speed, typhoon moving speed, and typhoon path length; among them, the typhoon best track dataset uses data publicly obtained from meteorological agencies.

3. The method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints according to claim 2, wherein Step 2 includes: Step 2.1, calculate the buffer polygon P with width L for the target typhoon path; Step 2.2, traverse all the typhoons to be identified in the historical typhoon database, and count the percentage of the typhoon centers of each typhoon to be identified that are located within the polygon P Among them, N in is the number of typhoon centers located inside the polygon P in a typhoon path to be recognized, and N total is the total number of typhoon centers in the typhoon path to be recognized; Step 2.3, when the path length ratio between the target typhoon and the typhoon to be identified satisfies ratio len ≥R1, ratio len ≤1 / R1, and ratio loc ≥R2, it is preliminarily identified that the typhoon path positions are similar and included in the subset of typhoons to be identified; Among them, R1 represents the similarity threshold of typhoon path length, and R2 represents the percentage threshold of the typhoon center positions on the typhoon path to be identified falling into the polygon P.

4. A similarity typhoon identification method based on vector-raster analysis with multi-parameter constraints according to claim 3, characterized in that In Step 4, the typhoon path similarity index S constrained by different combinations of multiple parameters such as typhoon path position, intensity, and / or moving speed is defined in the following form: Among them, d i is the closest distance between the position of any typhoon center of the target typhoon and the path of the typhoon to be identified, with the unit of Km; D start represents the spherical distance between the starting point of the path of the typhoon to be identified and the path of the target typhoon, and D end represents the spherical distance between the ending point of the path of the typhoon to be identified and the path of the target typhoon, with the unit of Km; () represents whether the parameter constraint effect can be selectively applied or not. The contents in the brackets in the formula are the typhoon intensity constraint and the traveling speed constraint in sequence. A is the weight coefficient of the traveling speed term, and the remaining three terms are all constraints on the typhoon path position; The smaller the value of this index, the higher the degree of simultaneous matching of the typhoon paths in terms of position, intensity, and / or moving speed; when this index is 0, the two typhoon paths are exactly the same; sorting according to the values of the identification index from small to large can identify typhoons with decreasing similarity in typhoon position, intensity, and / or moving speed.

5. A method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints according to claim 4, characterized in that Calculating the nearest position and the nearest distance corresponding to each typhoon center position on the target typhoon path in step 3 includes: interpolating and encrypting the typhoon path to be recognized, and for each typhoon center position on the target typhoon path, traversing each interpolated position on the typhoon to be recognized to obtain the position J corresponding to the minimum spherical distance i which is the nearest position of the typhoon center, and the corresponding distance is the nearest distance.

6. A method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints according to claim 5, characterized in that The calculation of the nearest distance corresponding to each typhoon center position on the target typhoon path in Step 3 includes: interpolating and densifying the typhoon path to be identified, and covering the target typhoon path with a rectangular area; for the interpolated and densified typhoon path to be identified, calculate the rasterized proximity within the rectangular area covered by the target typhoon path, and then the nearest distance at each center position of the target typhoon can be directly obtained by interpolating from the proximity field.

7. A method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints according to claim 5 or 6, characterized in that The formation of the pairing relationship for all typhoon center positions of the target typhoon in Step 3 is based on identifying its nearest position, including finding the nearest points on the interpolated and densified typhoon path to be identified for each center position on the target typhoon path, ensuring that the number of typhoon center pairing groups depends on the target typhoon, so as to ensure consistency in the index calculation of each typhoon to be identified.

8. A method for identifying similar typhoons based on vector-raster analysis with multi-parameter constraints according to claim 7, characterized in that, The typhoon path length in Step 1, as well as the spherical distance calculations in Steps 3 and 4, all use the great circle distance calculation.

Images