Height field and wind field mutual smoothing method based on physical rule constraint

By smoothing the bidirectional coupling relationship between the altitude field and the wind field, the problem of mismatch between wind direction and wind speed information in the wind field and altitude field is solved, thereby improving the physical consistency and analytical value of contour line drawing.

CN121479113APending Publication Date: 2026-02-06NANJING UNIV OF INFORMATION SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511608099.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-05
Publication Date
2026-02-06

AI Technical Summary

Technical Problem

Existing technologies rely on unidirectional or independent mathematical calculations when generating isolines in the geopotential height field, leading to a mismatch between wind direction and speed information in the wind field and the geopotential height field. This violates the laws of atmospheric dynamics and misleads weather analysis.

Method used

By preprocessing the two-dimensional height field and the wind field grid matrix, wind speed and wind direction are obtained, and mutual smoothing is performed based on the two-way coupling relationship. Anisotropic Gaussian kernels are used to smooth the gradient direction to ensure the physical consistency between the wind field and the height field.

Benefits of technology

This achieves unified coordination between the geopotential height field and the wind field, improves the quality of contour line drawing, ensures that contour lines are drawn along the wind direction, and enhances the analytical value of weather maps.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121479113A_ABST
    Figure CN121479113A_ABST
Patent Text Reader

Abstract

The invention is suitable for the technical field of meteorological data processing, and provides a height field and wind field mutual smoothing method based on physical rule constraints, which comprises the following steps: preprocessing a two-dimensional height field lattice point matrix hgtgrid, a two-dimensional wind field horizontal component lattice point matrix ugrid and a two-dimensional wind field vertical component lattice point matrix vgrid to obtain a wind speed and a wind direction, and performing mutual smoothing processing based on a bidirectional coupling relationship between the height field and the wind field, and outputting the height field and the wind field which are subjected to mutual smoothing when the number of cycles reaches a preset maximum number of iterations maxter. Therefore, by establishing a bidirectional dynamic coupling mechanism of the wind field and the height field, visual drawing of the contour line of the weather map is supported, and the method has the advantages of being high in physical consistency and overall coordination.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of meteorological data processing, and particularly relates to a height field and wind field mutual smoothing method based on physical rule constraint. BACKGROUND

[0002] In the field of meteorological data visualization, the drawing quality of the height field contour is crucial for meteorological analysts to quickly and accurately identify weather systems (such as high pressure, low pressure, trough line, ridge line), and clear, smooth and physically consistent contour lines can significantly improve the efficiency and accuracy of weather map interpretation. However, one of the core challenges in generating high-quality contour lines is to ensure their physical consistency with wind field data (e.g., conforming to geostrophic balance or quasi-geostrophic theory). Wind field and height field (wind pressure field) are not independent, but are physically coupled. However, in commonly used meteorological mapping software such as GrADS and NCL, the optimization of height field is mainly achieved by one-way or independent mathematical calculation (such as Gaussian smoothing, Barnes objective analysis, etc.) on wind field and height field.

[0003] However, the above method can suppress small-scale noise to some extent, but it cannot build overall physical coordination across scales between wind pressure fields, such as wind direction and contour line orientation consistency, and wind speed related to contour line spacing, resulting in mismatch between the corrected height field contour (isohypse) and wind direction and wind speed information of the wind field, which leads to errors in atmospheric dynamics and misleads weather analysis. SUMMARY

[0004] The embodiment of the application provides a height field and wind field mutual smoothing method based on physical rule constraint, which can solve the problem that the weather map isohypse and wind direction and wind speed information of the wind field are mismatched due to the one-way or independent mathematical calculation on the wind field and the height field in the weather analysis, and further produces errors in atmospheric dynamics and misleads weather analysis.

[0005] The embodiment of the application provides a height field and wind field mutual smoothing method based on physical rule constraint, which comprises: S1, pre-processing a two-dimensional height field grid point matrix hgt_grid, a two-dimensional wind field horizontal component grid point matrix u_grid and a two-dimensional wind field vertical component grid point matrix v_grid to obtain wind speed and wind direction; S2, performing mutual smoothing processing based on the bidirectional coupling relationship of the height field and the wind field, and outputting the mutually smoothed height field and wind field when the number of cycles reaches the preset maximum iteration number max_iter.

[0006] In a possible implementation manner of the first aspect, the two-dimensional height field grid point matrix hgt_grid, the two-dimensional wind field horizontal component grid point matrix u_grid, and the two-dimensional wind field vertical component grid point matrix v_grid all have a dimension of M rows and N columns; the grid points of u_grid and v_grid correspond to the same longitude and latitude as hgt_grid.

[0007] Optionally, in another possible implementation manner of the first aspect, the S1, preprocessing the two-dimensional height field grid point matrix hgt_grid, the two-dimensional wind field horizontal component grid point matrix u_grid, and the two-dimensional wind field vertical component grid point matrix v_grid to obtain the wind speed and the wind direction, includes:

[0008] Recording the data range of hgt_grid, u_grid, and v_grid:

[0009] h_min = min(hgt_grid), h_max = max(hgt_grid)

[0010] u_min = min(u_grid), u_max = max(u_grid)

[0011] v_min = min(v_grid), v_max = max(v_grid)

[0012] Converting the two-dimensional wind field into the wind speed and the wind direction:

[0013] The wind speed is

[0014] The wind direction is

[0015] Wherein, arctan2 is a complete mapping function of coordinates to radian angles, and is realized by two parameters to cover the whole quadrant (-π, π] and to process the boundary, mod represents the remainder, and the radian is mapped to [0, 2π).

[0016] Optionally, in another possible implementation manner of the first aspect, the S2, performing mutual smoothing processing based on the bidirectional coupling relationship between the height field and the wind field, includes:

[0017] S21, determining the wind field smoothing parameter by calculating the gradient strength and the gradient direction of the height field, and performing anisotropic Gaussian smoothing on the wind field with the gradient vertical direction as the main axis;

[0018] S22, determining the height field smoothing parameter according to the wind speed and the wind direction of the processed new wind field, and performing smoothing on the height field by using an anisotropic Gaussian kernel with the wind direction as the main axis.

[0019] Optionally, in another possible implementation of the first aspect, the S21, the wind field smoothing parameter is determined by calculating the gradient strength and gradient direction of the height field, and the wind field is anisotropically Gaussian smoothed with the gradient vertical direction as the main axis, including:

[0020] S211, calculating the gradient strength and gradient direction of the height field:

[0021]

[0022]

[0023]

[0024]

[0025] where i={1,…M}, j={1,…N}, and respectively represent the longitude grid distance and the latitude grid distance, is the gradient strength of the height field, is the gradient direction of the height field;

[0026] S212, calculating the wind field smoothing parameter:

[0027] The gradient strength factor is , and the value range is [0, 1]

[0028] The gradient vertical direction is , and the value range is [0, 2π);

[0029] S213, smoothing the wind speed by using the isotropic Gaussian kernel:

[0030] Calculate the smoothing strength

[0031] where sigma_base is the basic smoothing strength coefficient;

[0032] Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed wind speed:

[0033]

[0034] where i={1,…,M} and j={1,…,N}, and the Gaussian weight of each grid point is , The dimension of is [k_size, k_size], k_size is the size of the Gaussian kernel, ki and kj are the Gaussian kernel generation seeds of k_size=5, ki=kj={-2, -1, 0, 1, 2};

[0035] S214, smoothing the wind direction using an anisotropic Gaussian kernel:

[0036] Compute gradient vertical direction Smooth strength ;

[0037] Compute gradient direction smooth strength ;

[0038] Construct a rotation Gaussian kernel with the principal axis direction, and calculate the Gaussian weight of each grid point:

[0039]

[0040] where i={1,…,M}, j={1,…,N}, is the Gaussian weight of each grid point, with dimensions [k_size,k_size], k_size is the size of the Gaussian kernel, ki, kj are Gaussian kernel generation seeds of size k_size=5 ki=kj={-2,-1,0,1,2};

[0041] Apply Gaussian smoothing to each grid point (i,j) to get the corresponding smoothed wind direction :

[0042]

[0043]

[0044]

[0045] where i={1,…,M}, j={1,…,N}, is the smoothed wind direction horizontal component, is the smoothed wind direction vertical component;

[0046] S215, by converting the smoothed wind speed and direction into wind field component form, and performing numerical range normalization and inverse normalization, the new wind field variables are calculated and :

[0047]

[0048]

[0049]

[0050]

[0051]

[0052]

[0053] in, These are the smoothed horizontal and vertical components of the wind field. The horizontal and vertical components of the wind field after normalization and denormalization of the numerical range. This refers to the wind speed and direction of the new air field after the change.

[0054] Optionally, in another possible implementation of the first aspect, S22 above, determining the height field smoothing parameters based on the wind speed and direction of the processed new wind field, and smoothing the height field using an anisotropic Gaussian kernel with the wind direction as the main axis, includes:

[0055] S221. Calculate the height field smoothing parameters:

[0056] Wind speed intensity factor is Its range is [0,1];

[0057] S222. Smooth the height field using anisotropic Gaussian kernels:

[0058] Calculate wind direction Smoothness strength ;

[0059] Calculate the smoothing intensity in the vertical direction of the wind. ;

[0060] by Construct a rotating Gaussian kernel along the main axis and calculate the Gaussian weight for each grid point:

[0061]

[0062] Where i = {1, ..., M}, j = {1, ..., N} The Gaussian weights for each grid point are defined in dimension [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5: ki=kj={-2, -1, 0, 1, 2}.

[0063] Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed height field:

[0064]

[0065] Where i = {1, ..., M}, j = {1, ..., N};

[0066] S223. Perform numerical range normalization and denormalization on the smoothed height field:

[0067]

[0068] in, The height field is the result of normalization and denormalization of the numerical range.

[0069] Optionally, in another possible implementation of the first aspect, before outputting the mutually smoothed height field and wind field when the number of iterations reaches the preset maximum number of iterations max_iter, the following is also included:

[0070] Calculate the difference in field variables between the current iteration and the previous iteration:

[0071]

[0072]

[0073]

[0074]

[0075] in, This represents the maximum difference in the horizontal components of the two-dimensional wind field. This represents the maximum difference between the vertical components of the two-dimensional wind field. This represents the maximum difference in the two-dimensional height field. It is the maximum value of the first three;

[0076] Update wind field and altitude field:

[0077] Update u_grid with u_scaled, update v_grid with v_scaled, update w_spe with ws_new, update w_dir with wd_new, and update hgt_grid with hgt_scaled;

[0078] when If the iteration is terminated early, or the maximum number of iterations is reached, the updated u_grid, v_grid, and hgt_grid will be output as the smoothed wind field and height field. This is the convergence threshold.

[0079] Beneficial effects: The proposed solution first preprocesses the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid to obtain wind speed and wind direction. Then, based on the bidirectional coupling relationship between the height field and the wind field, mutual smoothing is performed. When the number of iterations reaches the preset maximum number of iterations max_iter, the mutually smoothed height field and wind field are output. This application does not process the geopotential height field or wind field independently or unidirectionally. Instead, it introduces a coupling constraint term based on physical rules to interactively guide the geopotential height field and wind field to adjust towards each other's states during the smoothing process. This enhances the unified coordination between wind field and geopotential height field data, thereby improving the quality of subsequent contour line drawing. This application introduces a constraint mechanism in two-dimensional space: the vertical direction of the geopotential height field gradient is unified with the opposite directions of the wind direction, forming a strong smoothing direction for the anisotropic Gaussian kernel, thus ensuring that contour lines are drawn along the direction of wind travel. Furthermore, the geopotential height field gradient intensity is unified with the wind speed intensity; lower intensity results in stronger smoothing with the isotropic Gaussian kernel, thus achieving a relatively larger contour line spacing in areas with low wind speeds. Therefore, this application is not limited to single weather scales such as typhoons or gradient winds, but can comprehensively improve the contour line drawing effect from large-scale circulation to small-to-medium-scale systems, significantly enhancing the analytical value of visualization products. Attached Figure Description

[0080] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0081] Figure 1 This is a flowchart illustrating a method for smoothing the height field and wind field based on physical rule constraints, provided in an embodiment of this application.

[0082] Figure 2 This is a schematic diagram of 500hPa upper-air weather after Gaussian smoothing of the altitude field, provided in one embodiment of this application;

[0083] Figure 3 This is a schematic diagram of 500hPa upper-air weather with unidirectional geostrophic balance weighted correction from wind field to height field provided in an embodiment of this application;

[0084] Figure 4 This is a schematic diagram of 500hPa upper-air weather with smoothed wind and altitude fields provided in one embodiment of this application;

[0085] Figure 5 This is a schematic diagram of 500hPa upper-air contour lines provided by artificial analysis in one embodiment of this application. Detailed Implementation

[0086] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0087] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0088] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0089] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0090] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0091] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0092] The following is a detailed description of a method for smoothing the height field and wind field based on physical rule constraints provided in this application, with reference to the accompanying drawings.

[0093] Figure 1 The diagram shows a flowchart of a method for smoothing the height field and wind field based on physical rule constraints, as provided in an embodiment of this application.

[0094] like Figure 1 As shown, the method for smoothing the height field and wind field based on physical rule constraints includes the following steps:

[0095] S1. Preprocess the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid to obtain wind speed and wind direction.

[0096] It should be noted that the dimensions of the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid are all M rows and N columns; the latitude and longitude of the grid points in u_grid and v_grid correspond to those in hgt_grid.

[0097] Furthermore, in this embodiment of the application, step S1 includes:

[0098] Record the data ranges of hgt_grid, u_grid, and v_grid:

[0099] h_min=min(hgt_grid),h_max=max(hgt_grid)

[0100] u_min=min(u_grid),u_max=max(u_grid)

[0101] v_min=min(v_grid), v_max=max(v_grid);

[0102] Convert the two-dimensional wind field into wind speed and wind direction:

[0103] Wind speed is

[0104] Wind direction is

[0105] Where arctan2 is a complete mapping function from coordinates to radian angles. It achieves full quadrant coverage and boundary handling capability through two parameters (-π,π], and mod represents the modulo operation, mapping radians to [0,2π].

[0106] S2. Based on the bidirectional coupling relationship between the altitude field and the wind field, perform mutual smoothing, and when the number of iterations reaches the preset maximum number of iterations max_iter, output the mutually smoothed altitude field and wind field.

[0107] For example, the default value for the maximum number of iterations, max_iter, is 3.

[0108] Furthermore, in this embodiment of the application, step S2 includes:

[0109] S21. Determine the wind field smoothing parameters by calculating the gradient intensity and gradient direction of the height field, and perform anisotropic Gaussian smoothing on the wind field with the gradient vertical direction as the main axis.

[0110] S22. Determine the height field smoothing parameters based on the wind speed and direction of the processed fresh air field, and smooth the height field using an anisotropic Gaussian kernel with the wind direction as the main axis.

[0111] Furthermore, in this embodiment of the application, step S21 includes:

[0112] S211. Calculate the gradient strength and gradient direction of the height field:

[0113]

[0114]

[0115]

[0116]

[0117] Where i = {1, ..., M}, j = {1, ..., N}, and These represent the longitude grid distance and the latitude grid distance, respectively. It should be noted that if the dimension exceeds the limit after adding or subtracting the index from the boundary point, the data of that boundary point is taken and divided only by the single grid distance. The gradient intensity of the height field, The gradient direction of the height field;

[0118] S212. Calculate wind field smoothing parameters:

[0119] Gradient strength factor is Its range is [0,1].

[0120] The gradient is perpendicular to the direction of the gradient. Its range is [0, 2π);

[0121] S213. Smoothing wind speed using an isotropic Gaussian kernel:

[0122] Calculate smoothing strength

[0123] Among them, sigma_base is the basic smoothing intensity coefficient (default value is 1.0);

[0124] Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed wind speed:

[0125]

[0126] Where i = {1, ..., M}, j = {1, ..., N}, and the Gaussian weight of each grid point is... , The dimension is [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5, ki=kj={-2, -1, 0, 1, 2}. It should be noted that if the dimension exceeds the limit after adding or subtracting the index at the boundary point, the out-of-bounds data and the corresponding Gaussian weights are ignored and not included in the summation.

[0127] S214. Smoothing wind direction using anisotropic Gaussian kernels:

[0128] Calculate the gradient vertical direction Smoothing intensity (strong smoothing) ;

[0129] Calculate the gradient direction smoothing strength (weak smoothing) ;

[0130] by Construct a rotating Gaussian kernel along the main axis and calculate the Gaussian weight for each grid point:

[0131]

[0132] Where i = {1, ..., M}, j = {1, ..., N}, The Gaussian weights for each grid point are defined in dimension [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5: ki=kj={-2, -1, 0, 1, 2}.

[0133] Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed wind direction. :

[0134]

[0135]

[0136]

[0137] Where i = {1, ..., M}, j = {1, ..., N}, This represents the smoothed horizontal component of the wind direction. The vertical component of the wind direction is smoothed. It should be noted that if the dimension exceeds the limit after adding or subtracting the index at the boundary point, the out-of-bounds data and the corresponding Gaussian weight are ignored and not included in the summation.

[0138] S215. By converting the smoothed wind speed and direction back into wind field component form, and performing numerical range normalization and denormalization, new wind field variables are calculated. and :

[0139]

[0140]

[0141]

[0142]

[0143]

[0144]

[0145] in, These are the smoothed horizontal and vertical components of the wind field. The horizontal and vertical components of the wind field after normalization and denormalization of the numerical range. This refers to the wind speed and direction of the new air field after the change.

[0146] Furthermore, in this embodiment of the application, step S22 includes:

[0147] S221. Calculate the height field smoothing parameters:

[0148] Wind speed intensity factor is Its range is [0,1];

[0149] S222. Smooth the height field using anisotropic Gaussian kernels:

[0150] Calculate wind direction Smoothing intensity (strong smoothing) ;

[0151] Calculate the smoothing intensity (weak smoothing) in the vertical direction of the wind. ;

[0152] by Construct a rotating Gaussian kernel along the main axis and calculate the Gaussian weight for each grid point:

[0153]

[0154] Where i = {1, ..., M}, j = {1, ..., N} The Gaussian weights for each grid point are defined in dimension [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5: ki=kj={-2, -1, 0, 1, 2}.

[0155] Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed height field:

[0156]

[0157] Where i={1,…,M}, j={1,…,N}, it should be noted that if the dimension exceeds the limit after adding or subtracting the index at the boundary point, the out-of-bounds data and the corresponding Gaussian weight are ignored and not included in the summation;

[0158] S223. Perform numerical range normalization and denormalization on the smoothed height field:

[0159]

[0160] in, The height field is the result of normalization and denormalization of the numerical range.

[0161] Furthermore, in this embodiment of the application, before outputting the mutually smoothed height field and wind field when the number of iterations reaches the preset maximum number of iterations max_iter, the method further includes:

[0162] Calculate the difference in field variables between the current iteration and the previous iteration:

[0163]

[0164]

[0165]

[0166]

[0167] in, This represents the maximum difference in the horizontal components of the two-dimensional wind field. This represents the maximum difference between the vertical components of the two-dimensional wind field. This represents the maximum difference in the two-dimensional height field. It is the maximum value of the first three;

[0168] Update wind field and altitude field:

[0169] Update u_grid with u_scaled, update v_grid with v_scaled, update w_spe with ws_new, update w_dir with wd_new, and update hgt_grid with hgt_scaled;

[0170] when If the iteration is terminated early, or the maximum number of iterations is reached, the updated u_grid, v_grid, and hgt_grid will be output as the smoothed wind field and height field. This is the convergence threshold (default value is 0.1).

[0171] This application provides a method for smoothing the height field and wind field based on physical rule constraints. The method preprocesses the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid to obtain wind speed and direction. Then, based on the bidirectional coupling relationship between the height field and wind field, mutual smoothing is performed. When the number of iterations reaches a preset maximum number of iterations max_iter, the smoothed height field and wind field are output. Thus, by establishing a bidirectional dynamic coupling mechanism between the wind field and the height field, it supports the visualization of contour lines on weather maps, exhibiting strong physical consistency and high overall coordination.

[0172] The following specific embodiment illustrates the solution of this application.

[0173] Optionally, an example of the contour line drawing process at 500 hPa altitude and the analysis of the drawing results can be provided. The input data is diamond4 (height field) and diamond11 (wind field UV component field) format data issued by the National Meteorological Administration, and the specific formats are shown in Table 1 and Table 2.

[0174] Table 1

[0175]

[0176] Note: Height is fixed at 0.

[0177] Table 2

[0178]

[0179] Read the above data and organize it into a height field grid matrix hgt_grid, a wind field horizontal component grid matrix u_grid, and a wind field vertical component grid matrix v_grid, respectively, and obtain the latitude and longitude grid spacing as follows: , Then, using the height field and wind field smoothing method proposed in this application, smoothed hgt_grid, u_grid, and v_grid are obtained through iteration; finally, hgt_grid is used to draw contour lines, and u_grid and v_grid are used to draw wind arrow and wind speed color maps, and the drawing results are compared and analyzed.

[0180] from Figures 2-5 It can be seen that, regarding the influence of wind direction on contour lines, only the altitude field undergoes Gaussian smoothing (e.g., Figure 2 A low-pressure area of ​​572 dagpm is retained in the lower left corner, but this area does not conform to the gradient wind principle because the wind direction at the closed contour lines of the low-pressure area does not exhibit the same closed unidirectional rotation; instead, an opening appears on its upper right side. Therefore, the contour lines at this location are drawn incorrectly, and this error cannot be corrected without considering the wind field. However, wind field correction based on geostrophic equilibrium, which unidirectionally corrects the height field (e.g., ... Figure 3 The method in this application removes the closed contour line at 572 dagpm, which is incorrect because the open contour line at 572 dagpm adjacent to its upper right should also be drawn. Therefore, the method in this application (such as...) Figure 4 By performing multiple iterations of strong smoothing on the height field along the wind direction, the 572 dagpm closed contour line is integrated into the adjacent 572 dagpm open contour line to its upper right. This not only ensures that the same geopotential height values ​​can be completely plotted, but also conforms to the physical rule that contour lines are plotted along the direction of wind. Therefore, it is consistent with the results of manual analysis (such as...). Figure 5 ) is the most suitable.

[0181] Furthermore, regarding the influence of wind speed on contour lines, in order to comply with the physical rule that the spacing between contour lines is relatively larger in areas with low wind speeds, the method of this application (such as...) Figure 4 )contrast Figure 2 In low-wind-speed areas, such as the regions at latitudes and longitudes of 80-90°E and 50-60°N (as can be seen from the shaded map of this region, wind speeds are mainly below 8 m / s), the contour lines 548 and 552 dagpm shift slightly upwards, thus increasing the spacing. As for... Figure 3 The height field value is too high due to the correction, which differs from the results of manual analysis (such as...). Figure 5 This resulted in a significant deviation.

[0182] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0183] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for smoothing the mutual interaction between altitude field and wind field based on physical rule constraints, characterized in that, include: S1. Preprocess the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid to obtain wind speed and wind direction. S2. Based on the bidirectional coupling relationship between the altitude field and the wind field, perform mutual smoothing, and when the number of iterations reaches the preset maximum number of iterations max_iter, output the mutually smoothed altitude field and wind field.

2. The method as described in claim 1, characterized in that, The dimensions of the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid are all M rows and N columns; the latitude and longitude of the grid points of u_grid and v_grid correspond to those of hgt_grid.

3. The method as described in claim 2, characterized in that, S1 involves preprocessing the two-dimensional height field grid matrix hgt_grid, the two-dimensional wind field horizontal component grid matrix u_grid, and the two-dimensional wind field vertical component grid matrix v_grid to obtain wind speed and direction, including: Record the data ranges of the hgt_grid, u_grid, and v_grid: h_min=min(hgt_grid),h_max=max(hgt_grid) u_min=min(u_grid),u_max=max(u_grid) v_min=min(v_grid), v_max=max(v_grid); Convert the two-dimensional wind field into wind speed and wind direction: Wind speed is ; Wind direction is ; Where arctan2 is a complete mapping function from coordinates to radian angles. It achieves full quadrant coverage and boundary handling capability through two parameters (-π,π], and mod represents the modulo operation, which maps radians to [0,2π].

4. The method as described in claim 3, characterized in that, S2, based on the bidirectional coupling relationship between the altitude field and the wind field, performs mutual smoothing processing, including: S21. Determine the wind field smoothing parameters by calculating the gradient intensity and gradient direction of the height field, and perform anisotropic Gaussian smoothing on the wind field with the gradient vertical direction as the main axis. S22. Determine the height field smoothing parameters based on the wind speed and direction of the processed fresh air field, and smooth the height field using an anisotropic Gaussian kernel with the wind direction as the main axis.

5. The method as described in claim 4, characterized in that, S21, determining wind field smoothing parameters by calculating the gradient intensity and gradient direction of the height field, and performing anisotropic Gaussian smoothing of the wind field with the gradient perpendicular direction as the main axis, includes: S211. Calculate the gradient strength and gradient direction of the height field: ; ; ; ; Where i = {1, ..., M}, j = {1, ..., N}, and These represent longitude grid spacing and latitude grid spacing, respectively. The gradient intensity of the height field, The gradient direction of the height field; S212. Calculate wind field smoothing parameters: Gradient strength factor is Its range is [0,1]. The gradient is perpendicular to the direction of the gradient. Its range is [0, 2π); S213. Smoothing wind speed using an isotropic Gaussian kernel: Calculate smoothing strength ; Among them, sigma_base is the basic smoothing strength coefficient; Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed wind speed: ; Where i = {1, ..., M}, j = {1, ..., N}, and the Gaussian weight of each grid point is... , The dimension is [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are Gaussian kernel generation seeds of size k_size=5, ki=kj={-2, -1, 0, 1, 2}; S214. Smoothing wind direction using anisotropic Gaussian kernels: Calculate the gradient vertical direction Smoothness strength ; Calculate gradient direction smoothing intensity ; As stated Construct a rotating Gaussian kernel along the main axis and calculate the Gaussian weight for each grid point: ; Where i = {1, ..., M}, j = {1, ..., N}, The Gaussian weights for each grid point are defined in dimension [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5: ki=kj={-2, -1, 0, 1, 2}. Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed wind direction. : ; ; ; Where i = {1, ..., M}, j = {1, ..., N}, This represents the smoothed horizontal component of the wind direction. The smoothed vertical component of the wind direction; S215. By converting the smoothed wind speed and direction back into wind field component form, and performing numerical range normalization and denormalization, new wind field variables are calculated. and : ; ; ; ; ; ; in, These are the smoothed horizontal and vertical components of the wind field. The horizontal and vertical components of the wind field after normalization and denormalization of the numerical range. This refers to the wind speed and direction of the new air field after the change.

6. The method as described in claim 5, characterized in that, S22, determining the height field smoothing parameters based on the wind speed and direction of the processed fresh air field, and smoothing the height field using an anisotropic Gaussian kernel with the wind direction as the main axis, includes: S221. Calculate the height field smoothing parameters: Wind speed intensity factor is Its range is [0,1]; S222. Smooth the height field using anisotropic Gaussian kernels: Calculate wind direction Smoothness strength ; Calculate the smoothing intensity in the vertical direction of the wind. ; by Construct a rotating Gaussian kernel along the main axis and calculate the Gaussian weight for each grid point: ; Where i = {1, ..., M}, j = {1, ..., N} The Gaussian weights for each grid point are defined in dimension [k_size, k_size], where k_size is the size of the Gaussian kernel, and ki and kj are the Gaussian kernel generation seeds of size k_size=5: ki=kj={-2, -1, 0, 1, 2}. Apply Gaussian smoothing to each grid point (i, j) to obtain the corresponding smoothed height field: ; Where i = {1, ..., M}, j = {1, ..., N}; S223. Perform numerical range normalization and denormalization on the smoothed height field: ; in, The height field is the result of normalization and denormalization of the numerical range.

7. The method as described in claim 6, characterized in that, Before outputting the smoothed height field and wind field when the number of iterations reaches the preset maximum number of iterations max_iter, the process also includes: Calculate the difference in field variables between the current iteration and the previous iteration: ; ; ; ; in, This represents the maximum difference in the horizontal components of the two-dimensional wind field. This represents the maximum difference between the vertical components of the two-dimensional wind field. This represents the maximum difference in the two-dimensional height field. It is the maximum value of the first three; Update wind field and altitude field: Update u_grid with u_scaled, update v_grid with v_scaled, update w_spe with ws_new, update w_dir with wd_new, and update hgt_grid with hgt_scaled; when If the iteration is terminated early, or the maximum number of iterations is reached, the updated u_grid, v_grid, and hgt_grid will be output as the smoothed wind field and height field. This is the convergence threshold.