A method, device and system for determining an observation sensitive area with high influence weather forecast accuracy

Abstract

The application discloses a kind of methods for determining observation sensitive area affecting numerical weather prediction accuracy, device, system and computer readable storage medium, the method includes determining target area range and date;Obtain the value of CMA-GFS4DVAR (global four-dimensional variational data assimilation numerical prediction system) operational analysis field;The value of the operational analysis field is interpolated to the corresponding resolution of calculating singular vector SVs;The singular vector SVs is generated by calculation;According to the linear combination of the singular vector SVs, a sensitive area is generated;The display graph of the sensitive area is drawn.

Description

Technical Field

[0001] This invention belongs to the field of meteorological technology, specifically, it relates to a method, device and system for determining observation sensitive areas with high accuracy in weather forecasting. Background Technology

[0002] Currently, my country has built a world-class integrated three-dimensional meteorological observation network system that combines ground, air, and space. Nearly tens of thousands of automatic ground observation stations, hundreds of weather radars, and multiple Fengyun meteorological satellites closely monitor my country's weather and climate conditions. Furthermore, a meteorological service system with extensive coverage has been established, effectively serving dozens of departments, hundreds of industries, and hundreds of millions of people.

[0003] However, the existing observation network still has a sparse grid distribution compared to high-resolution numerical weather prediction, especially in regions with distinctive characteristics, such as the Tibetan Plateau, where observation points are even scarcer. This coarse resolution of the observation network has become one of the key factors affecting the accuracy of initial numerical weather prediction values, thus restricting the improvement of numerical weather prediction quality. Therefore, in order to improve the level of meteorological numerical weather prediction technology in key meteorological science and technology fields, improving and strengthening the construction and layout of the observation network is crucial to improving the development of numerical weather prediction operations.

[0004] Ideally, in high-resolution numerical weather prediction models, each grid point in the observation network would have high-precision observations nearby, but this is impossible in reality due to the enormous cost of such a high-precision observation network. To find the optimal observation layout configuration with limited resources, this invention proposes a method for determining observation sensitive areas that affect weather forecast accuracy. This method can be used during severe weather events (such as typhoons) and major weather service periods, or to provide intelligent observation layout suggestions for key observation areas with special needs (such as along the Sichuan-Tibet Railway).

[0005] In summary, the existing meteorological observation network system still has low accuracy in weather forecasting and cannot provide more accurate weather forecasts, especially for areas or service needs with high accuracy requirements.

[0006] Therefore, improving weather forecasting under existing conditions is an urgent problem to be solved. Summary of the Invention

[0007] To address the problems existing in the prior art, the present invention provides a method, apparatus, and system for determining observation sensitive areas that affect the accuracy of weather forecasts.

[0008] The present invention adopts the following technical solution:

[0009] Firstly, a method for determining observation-sensitive areas with high accuracy in weather forecasting includes:

[0010] Determine the target area and date;

[0011] Obtain the numerical values ​​of the CMA-GFS 4DVAR business analysis field;

[0012] The numerical values ​​of the business analysis field are interpolated to the corresponding resolution for calculating the singular vectors (SVs);

[0013] The singular vectors SVs are calculated and generated.

[0014] The sensitive region is generated by linear combination of the singular vectors SVs;

[0015] Draw a display diagram of the sensitive area.

[0016] Furthermore, the calculation and generation of the singular vectors SVs includes:

[0017] The first step is to convert the Eulerian vectors into their corresponding vectors using the Lanczos algorithm. Perform the inverse E of the total energy modulus transformation operator E -1 Transform it into an initial perturbation vector X = (u', v', (θ')', (Π')');

[0018] The second step involves performing a parallel CMA-GFS tangent mode backward integration (time t) on the initial perturbation vector x to obtain the evolved perturbation vector X. t =(u',v',(θ'′)',(Π')') t ;;

[0019] The third step is to examine X. t The total energy transformation operator E is performed to obtain the L for input into the CMA-GFS adjoint mode. T The input perturbation vector X t * =(u',v',(θ')',(Π')') t * ;

[0020] The fourth step is to process the accompanying perturbation vector X. t *Utilizing the parallel CMA-GFS adjoint mode L T Integrating forward over time interval t yields the adjoint gradient vector X0 at the initial time. * =(u',v',(θ')',(Π')')0 * ;

[0021] The calculation of singular vectors in CMA-GFS requires the CMA-GFS tangential linear model and its adjoint model. The forecast variables are the tangential linear perturbation variables of the horizontal wind component (u,v), perturbation potential temperature (θ'), and perturbation dimensionless pressure (Π') in the CMA-GFS model forecast variables, which can be expressed as (u',v',(θ')',(Π')').

[0022] Fifth step, for the adjoint gradient vector X0 * Implement E -1 The transformation yields Eulerian space vectors;

[0023] Step 6: Iterate through steps 1 to 5, and use the fastest growing Euler vector as the termination condition for the iteration. The fastest growing Euler vector is referred to as the singular vector SVs.

[0024] Furthermore, the Singular Vectors (SVs) perturbations are a set of orthogonal perturbations that grow the fastest within a certain time interval (optimization time) based on the total atmospheric energy weighted mode in the tangent linear model.

[0025] Furthermore, the formula adopted for the total energy mode transformation is:

[0026]

[0027] In the formula, L T For the adjoint mode of the CMA-GFS tangent linear mode, λ i For matrix operators (L T P T E T The singular value of EPL, X i (t0) is a matrix (L) T P T E T EPL) belongs to λ i The corresponding i-th singular vector, E is the weight modulus (normalized perturbation in energy units) that measures the magnitude of the perturbation.

[0028] Furthermore, the formula for calculating the total energy modulus E is as follows:

[0029]

[0030] In the formula, the sum of the first two terms represents the perturbation kinetic energy, and the latter two terms represent the perturbation potential energy. The third and fourth terms represent the contributions of the perturbation potential temperature (θ')' and the perturbation dimensionless pressure (Π')' components to the potential energy, respectively. In formula (3) in For terrain-following coordinates, λ and C represents the longitude and latitude of the model spherical coordinates, respectively. P T is the specific heat at constant pressure of dry air. r θ r 、∏ r and ρ r These represent the reference temperature, reference potential temperature, reference dimensionless pressure, and reference density, respectively, and their calculation formulas are as follows: R dp is the gas constant for dry air. r For standard atmosphere, T r Here is the reference temperature, and g is the acceleration due to gravity.

[0031] Furthermore, the formula for calculating the sensitive area is: In the formula, λ i V is the i-th singular value calculated in Formula 2. i Let i be the i-th singular vector.

[0032] Furthermore, the sensitive area is displayed as a percentage graph.

[0033] Secondly, a device for determining observation-sensitive areas with high accuracy in weather forecasting includes:

[0034] The determination module is used to determine the target area range and date;

[0035] The acquisition module is used to acquire values ​​from the CMA-GFS business analysis field;

[0036] The configuration module is used to interpolate the numerical values ​​of the business analysis field to the corresponding resolution for calculating singular vectors (SVs).

[0037] The calculation module is used to calculate and generate the singular vectors SVs;

[0038] The sensitive region is generated by linear combination of the singular vectors SVs;

[0039] The drawing module is used to draw a display diagram of the sensitive area.

[0040] Thirdly, a system for determining observation-sensitive areas with high accuracy in weather forecasting, the system comprising:

[0041] One or more processors;

[0042] Storage device for storing one or more programs;

[0043] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-described method for determining observation sensitive areas that affect the accuracy of weather forecasts.

[0044] Fourthly, a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for determining observation sensitive areas with high accuracy in weather forecasting.

[0045] Beneficial effects

[0046] Compared to existing technologies, the advantages of this invention are as follows: It provides a method for determining observation sensitive areas that affect the accuracy of weather forecasts, including: determining the target area range and date; acquiring numerical values ​​from the CMA-GFS operational analysis field; interpolating the numerical values ​​of the operational analysis field to the corresponding resolution for calculating singular vectors (SVs); calculating and generating the singular vectors (SVs); generating sensitive areas based on linear combinations of the singular vectors (SVs); and plotting a display map of the sensitive areas. This method can improve the accuracy of weather forecasts based on existing meteorological observation network systems, providing high-precision weather forecasts for high-impact weather systems and major weather services. Attached Figure Description

[0047] Figure 1 This is a flowchart of a method for determining observation sensitive areas with high accuracy in weather forecasting according to the present invention;

[0048] Figure 2 This is a specific experimental configuration of the global forecast model version CMA-GFSV3.2 used in one embodiment of the present invention for the forecast sensitivity experiment;

[0049] Figure 3 The structural calculation flow of the CMA-GFS mode program in one embodiment of the present invention;

[0050] Figures 4(A) and (B) are respectively a flowchart and a parameter relationship diagram of step S4 for calculating and generating the singular vectors SVs in one embodiment;

[0051] Figure 5 This is a structural diagram of a device for determining observation sensitive areas with high accuracy in weather forecasting according to the present invention;

[0052] Figure 6 This is a structural diagram of an electronic equipment system provided for an embodiment of the operating environment of a method for determining observation sensitive areas with high accuracy in weather forecasting according to the present invention;

[0053] Figure 7 A structural diagram of a computer-readable storage medium provided for an embodiment of the operating environment of a method for determining observation sensitive areas with high accuracy in weather forecasting according to the present invention;

[0054] Figure 8 This is a map showing the setting of sensitive areas for severe weather forecasting along the Sichuan-Tibet Railway in one embodiment of the present invention;

[0055] Figure 9 This is a map showing the annual average level distribution of forecast sensitive areas affecting the Sichuan-Tibet Railway, according to an embodiment of the present invention.

[0056] Figure 10This is an embodiment of the present invention showing the annual average vertical distribution of forecast sensitive areas affecting the region along the Sichuan-Tibet Railway. Detailed Implementation

[0057] The present invention will be further described below with reference to specific embodiments.

[0058] The terms “comprising” and “including” as used herein should be understood as inclusive and open-ended, not exclusive. Specifically, when the terms “comprising” and “including” and their synonyms are used in the specification and claims, they mean that the specified features, steps, or components are included. These terms should not be construed as excluding the existence of other features, steps, or components.

[0059] This invention provides a method for determining observation-sensitive areas with high accuracy in weather forecasting. For example... Figure 1 As shown, the method includes the following steps:

[0060] S1: Determine the target area and date;

[0061] S2: Obtain the numerical values ​​of the business analysis field in CMA-GFS.

[0062] exist Figure 2 The diagram shows the setup for the forecast sensitivity experiment. Specifically, Figure 2 In one embodiment of the present invention, the specific experimental configuration of CMA-GFS V3.2 used in the forecast sensitivity test is described.

[0063] Furthermore, in one embodiment, the structural calculation process of the CMA-GFS mode program is as follows: Figure 3 First, the configuration and initialization procedures for the global CMA-GFS tangent and adjoint modes are initiated. Then, the configuration information table of SVD (namelist.svd) is read in to determine the number of iterations and the size of the matrix in Lanczos. Subsequently, the energy mode operator E matrix is ​​calculated. Finally, the most crucial step in solving CMA-GFS SVs is performed: the cyclic iteration of Lanczos, repeatedly calling the CMA-GFS tangent and adjoint modes until the number of iterations set in namelist.svd is completed, obtaining all SVs that meet the required precision, and determining whether to calculate the evolutionary SVs.

[0064] S3: Interpolate the numerical values ​​of the business analysis field to the corresponding resolution for calculating the singular vectors SVs.

[0065] In one embodiment, the singular vectors (SVs) perturbations are a set of orthogonal perturbations that grow fastest within a certain time interval (optimization time) based on the total atmospheric energy weighted mode in the tangent linear model.

[0066] For an initial perturbation vector X0, the vector is processed by the Global-Regional Assimilation and Prediction System (CMA-GFS) tangent linear model (denoted as L, and the adjoint model denoted as L). T The evolutionary perturbation vector X can be obtained by forward integration over a certain period of time. t =Evolution of LX0. The solution of singular vectors in CMA-GFS can be reduced to the evolution of the perturbation vector X. t The problem of maximizing the ratio between the magnitude of the disturbance vector and the magnitude of the initial disturbance vector X0:

[0067]

[0068] In the formula, P is the target area projection operator, which will obtain the SVs that affect the target area forecast in the future, and E is the weight modulus (normalized perturbation in energy units) that measures the magnitude of the perturbation.

[0069] S4: Calculate and generate the singular vectors SVs.

[0070] Figures 4(A) and (B) are respectively a flowchart and a parameter relationship diagram of step S4 for calculating and generating the singular vectors SVs in one embodiment. As shown in Figure 4, in one embodiment, step S4 for calculating and generating the singular vectors SVs includes:

[0071] S41: According to Lanczos' algorithm, convert the Eulerian vectors... The inverse E⁻¹ of the total energy modulus transformation operator E is converted into the initial perturbation vector X = (u', v', (θ')', (Π')').

[0072] The formula used for the total energy mode transformation is as follows:

[0073]

[0074] In the formula, L T For the adjoint mode of the CMA-GFS tangent linear mode, λ i For matrix operators (L T P T E T The singular value of EPL, X i (t0) is a matrix (L) T P T E T EPL) belongs to λ i The corresponding i-th singular vector, E is the weight modulus (normalized perturbation in energy units) that measures the magnitude of the perturbation.

[0075] In one embodiment, the CMA-GFS singular vector computation problem defined by formula (2) can be solved using the Lanczos iterative algorithm. During the iteration process, it is necessary to integrate the atmospheric tangent linear mode L and the adjoint mode L multiple times. T Meanwhile, it can also be seen from formula (2) that the key factors determining the singular vector structure are mainly three aspects: 1) the selection of the weight modulus E to measure the magnitude of the singular vector perturbation; 2) the characteristics of the tangent linear and adjoint modes, mainly the use of the parameterization scheme of the linearized physical process; 3) the forward movement of the tangent linear mode L and the adjoint mode L T The length of the backward integration time (optimization time interval).

[0076] Furthermore, the formula for calculating the total energy modulus E is as follows:

[0077]

[0078] In the formula, the sum of the first two terms represents the perturbation kinetic energy, and the latter two terms represent the perturbation potential energy. The third and fourth terms represent the contributions of the perturbation potential temperature (θ')' and the perturbation dimensionless pressure (Π')' components to the potential energy, respectively. In formula (3) in For terrain-following coordinates, λ and C represents the longitude and latitude of the model spherical coordinates, respectively. P T is the specific heat at constant pressure of dry air. r θ r Π r and ρ r These represent the reference temperature, reference potential temperature, reference dimensionless pressure, and reference density, respectively, and their calculation formulas are as follows: R d p is the gas constant for dry air. r For standard atmosphere, T r Here is the reference temperature, and g is the acceleration due to gravity.

[0079] S42: Perform parallel CMA-GFS tangent mode backward integration on the initial perturbation vector x for time interval t to obtain the evolved perturbation vector X. t =(u',v',(θ')',(Π')') t ;

[0080] S43: In relation to x t Perform total energy transformation operator E transformation (E 2 ), to obtain the accompanying mode L for input to CMA-GFS. T Accompanying perturbation vector X t * =(u',v',(θ')',(Π')') t* ;

[0081] S44: For the accompanying perturbation vector x t *Utilizing the parallel CMA-GFS adjoint mode L T Integrating forward over time interval t yields the adjoint gradient vector X0 at the initial time. * =(u',v',(θ')',(Π')')0 * ;

[0082] Among them, the CMA-GFS tangent linear model and its adjoint model required for the CMA-GFS singular vector calculation, the forecast quantities of which are the horizontal wind components (u,v), perturbation potential temperature (θ'), and perturbation dimensionless pressure (Π') in the CMA-GFS model forecast variables can be expressed as: u', v', (θ')', and (Π')', respectively.

[0083] S45: Apply E to the adjoint gradient vector x0* -1 The transformation yields the Euler vector.

[0084] S46: Iteration steps one through five, obtaining the fastest growing Euler vector as the iteration termination condition, the fastest growing Euler vector being the singular vector SVs.

[0085] S5: Generate the sensitive region based on the linear combination of the singular vectors SVs;

[0086] The formula for calculating the sensitive area is as follows:

[0087] In the formula, λ i For the calculated i-th singular value, V i Let i be the i-th singular vector.

[0088] S6: Draw a display diagram of the sensitive area.

[0089] Furthermore, the sensitive area is displayed as a percentage graph.

[0090] In one embodiment, the method for determining observation sensitive areas affecting weather forecast accuracy of the present invention is applied to the sensitivity of severe weather forecasts along the Sichuan-Tibet Railway. Figure 8 This is a map showing the setting of sensitive areas for severe weather forecasting along the Sichuan-Tibet Railway in one embodiment of the present invention. Figure 8 In the diagram, the horizontal and vertical axes represent latitude and longitude.

[0091] like Figure 8As shown, the dashed boxes represent the target areas for severe weather forecasting along the Sichuan-Tibet Railway. Areas with a terrain elevation below 2500m are the source of the Yangtze River, areas with a terrain elevation above 4500m are the source of the Yellow River, and areas with a terrain elevation between 2500m and 4500m are the source of the Lancang River. In this embodiment, the sensitivity of the D01 area—the source of the three rivers (Yangtze River, Yellow River, and Lancang River) of the Asian Water Tower—to severe weather forecasting along the Sichuan-Tibet Railway is emphasized.

[0092] In a more detailed embodiment, such as Figure 9 This is an annual average distribution map of the forecast-sensitive areas affecting the Sichuan-Tibet Railway, according to an embodiment of the present invention. The statistical period for the one-year results of the weather forecast sensitivity test along the Sichuan-Tibet Railway is from December 1, 2018 to November 30, 2019. Figure 9 In the diagram, the horizontal and vertical axes represent latitude and longitude. The horizontal distribution of the annual average results of the sensitivity to forecasts along the Sichuan-Tibet Railway is shown. Figure 9 As can be seen in the image, areas with a gray value greater than 0.8 are forecast-sensitive areas. These areas are mainly distributed in the southern part of the Qinghai-Tibet Plateau and located in the northwestern part of the Sichuan-Tibet Railway. This is primarily because the westerly jet stream encounters the Qinghai-Tibet Plateau, is blocked in the western section of the plateau, splits into north and south branches, and then travels around the plateau before converging with the summer Indian monsoon from west of the plateau. This combined effect forms the forecast-sensitive areas affecting the 24-hour forecasts along the Sichuan-Tibet Railway. In the image, the forecast-sensitive areas (with gray values ​​greater than 0.8) basically cover the Three-River-Source region; therefore, the Three-River-Source region is one of the key sensitive areas affecting the forecasting effectiveness along the Sichuan-Tibet Railway.

[0093] Figure 10 This is an embodiment of the present invention showing the annual average vertical distribution of forecast-sensitive areas affecting the region along the Sichuan-Tibet Railway, where gray values ​​greater than 0.8 indicate forecast-sensitive areas. From Figure 10 As can be seen from the vertical distribution of the annual average sensitivity results affecting the forecasts along the Sichuan-Tibet Railway, the forecast-sensitive area (with a gray value greater than 0.8) has two large value centers in the vertical direction. The large value center located at the tropopause, i.e., at a pressure height of (100-150) hPa, reflects that the westerly jet stream begins to have a significant impact on the 24-hour forecasts of hazardous weather along the Sichuan-Tibet Railway as soon as it enters the tropopause of the Qinghai-Tibet Plateau. Figure 9 Annual average distribution of sensitive areas in the forecast Figure 1 As a result, the westerly winds that wind around the plateau and the summer Indian monsoon from the west of the plateau converge at the boundary of the Tibetan Plateau to the troposphere, where the air pressure is at (500-250) hPa. Under their combined influence, a high-value center of the sensitive area affecting the 24-hour forecast along the Sichuan-Tibet Highway is formed.

[0094] On the other hand, in one embodiment, the present invention provides a sensitive observation area determination device with high accuracy in weather forecasting, such as... Figure 5 As shown, the device includes:

[0095] The determination module is used to determine the target area range and date;

[0096] The acquisition module is used to acquire values ​​from the CMA-GFS 4DVAR business analysis field;

[0097] The configuration module is used to interpolate the numerical values ​​of the business analysis field to the corresponding resolution for calculating singular vectors (SVs).

[0098] The calculation module is used to calculate and generate the singular vectors SVs;

[0099] The sensitive region is generated by linear combination of the singular vectors SVs;

[0100] The drawing module is used to draw a display diagram of the sensitive area.

[0101] The method implemented by the device includes:

[0102] In one embodiment, the computation of generating the singular vectors SVs includes:

[0103] The first step is to convert the Eulerian vectors into their corresponding vectors using the Lanczos algorithm. Perform the inverse E of the total energy modulus transformation operator E -1 Transform it into an initial perturbation vector X = (u', v', (θ')', (Π')');

[0104] The second step involves performing a parallel CMA-GFS tangent mode backward integration (time t) on the initial perturbation vector x to obtain the evolved perturbation vector X. t =(u',v',(θ')',(Π')') t ;

[0105] The third step is to examine X. t The total energy transformation operator E is performed to obtain the L for input into the CMA-GFS adjoint mode. T The input perturbation vector X t *=(u',v',(θ')',(Π')') t * ;

[0106] The fourth step is to process the accompanying perturbation vector X. t * Using the parallel CMA-GFS adjoint mode L T Integrating forward over time interval t yields the adjoint gradient vector X0 at the initial time. *=(u',v',(θ')',(Π')')0 * ;

[0107] Among them, the CMA-GFS tangent linear model and its adjoint model required for the CMA-GFS singular vector calculation, whose forecast quantities are the horizontal wind components (u,v), perturbation potential temperature (θ'), and perturbation dimensionless pressure (Π') in the CMA-GFS model forecast variables can be expressed as u', v', (θ')', and (Π')', respectively.

[0108] Fifth step, for the adjoint gradient vector X0 * Implement E -1 The transformation yields Eulerian space vectors;

[0109] Step 6: Iterate through steps 1 to 5, and use the fastest growing Euler vector as the termination condition for the iteration. The fastest growing Euler vector is referred to as the singular vector SVs.

[0110] In one embodiment, the singular vectors (SVs) perturbations are a set of orthogonal perturbations that grow fastest within a certain time interval (optimization time) based on the total atmospheric energy weighted mode in the tangent linear model.

[0111] For an initial perturbation vector X0, after passing through the CMA-GFS tangent linear mode (denoted as L, and the adjoint mode denoted as L), T The evolutionary perturbation vector X can be obtained by forward integration over a certain period of time. t =Evolution of LX0. The solution of singular vectors in CMA-GFS can be reduced to the evolution of the perturbation vector X. t The problem of maximizing the ratio between the magnitude of the disturbance vector and the magnitude of the initial disturbance vector X0:

[0112]

[0113] In the formula, P is the target area projection operator, which will obtain the SVs that affect the target area forecast in the future, and E is the weight modulus (normalized perturbation in energy units) that measures the magnitude of the perturbation.

[0114] In one embodiment, the formula used for the total energy mode transformation is:

[0115]

[0116] In the formula, LT is the adjoint mode of the CMA-GFS tangent linear mode, and λi is the matrix operator (L T P T E T The i-th singular value of EPL, Xi(t0), is the matrix (L). T P T ET EPL) belongs to λ i The corresponding i-th singular vector, E is the weight modulus (normalized perturbation in energy units) that measures the magnitude of the perturbation.

[0117] In one embodiment, the formula for calculating the total energy modulus E is as follows:

[0118]

[0119] In the formula, the sum of the first two terms represents the perturbation kinetic energy, and the latter two terms represent the perturbation potential energy. The third and fourth terms represent the contributions of the perturbation potential temperature (θ')' and the perturbation dimensionless pressure (Π')' components to the potential energy, respectively. In formula (3) in For terrain-following coordinates, λ and C represents the longitude and latitude of the model spherical coordinates, respectively. P T is the specific heat at constant pressure of dry air. r θ r 、∏ r and ρ r These represent the reference temperature, reference potential temperature, reference dimensionless pressure, and reference density, respectively, and their calculation formulas are as follows: R d p is the gas constant for dry air. r For standard atmosphere, T r Here is the reference temperature, and g is the acceleration due to gravity.

[0120] In one embodiment, the formula for calculating the sensitive area is: In the formula, λ i V is the i-th singular value calculated in Formula 2. i Let i be the i-th singular vector.

[0121] In one embodiment, the sensitive area is displayed as a percentage graph.

[0122] On the other hand, in one embodiment, please refer to Figure 6 , Figure 6 This is a structural diagram of an electronic device provided in an embodiment of this application. Figure 6 As shown, the electronic device includes a processor, a memory, and a bus.

[0123] The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via a bus. When the machine-readable instructions are executed by the processor, they can perform the operations described above. Figure 1 The steps of the method for determining observation sensitive areas with high accuracy in weather forecasting shown in the method embodiment can be found in the method embodiment for specific implementation, and will not be repeated here.

[0124] On the other hand, in one embodiment, this application also provides a computer-readable storage medium, such as... Figure 7 As shown, the computer-readable storage medium stores a computer program that, when executed by a processor, can perform the above-described actions. Figure 1 The steps of the method for determining observation sensitive areas with high accuracy in weather forecasting shown in the method embodiment can be found in the method embodiment for specific implementation, and will not be repeated here.

[0125] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0126] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and other division methods may be used in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the functional units in the various embodiments of this application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0127] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0128] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the technical scope disclosed in this application. Such modifications, changes, 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 covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A method for determining observation-sensitive areas with high accuracy in weather forecasting, comprising: S1: Determine the target area and date; S2: Obtain the numerical values ​​of the business analysis field of CMA-GFS 4DVAR; S3: Interpolate the numerical values ​​of the business analysis field to the corresponding resolution for calculating the singular vectors SVs; S4: Calculate and generate the singular vectors SVs; Step S4 specifically includes: S41: According to Lanczos' algorithm, convert the Eulerian vectors... Perform total energy mode transformation operator The reverse Transform into initial perturbation vector ; S42: Perform parallel CMA-GFS tangent mode backward integration on the initial perturbation vector x over time interval t to obtain the evolved perturbation vector. ; S43: In response to Perform total energy transformation operator The transformation yields the accompanying mode used as input to the CMA-GFS. input perturbation vector ; S44: For the accompanying perturbation vector Using the parallel CMA-GFS adjoint mode Integrating forward by time interval t yields the adjoint gradient vector at the initial time. ; The calculation of singular vectors in CMA-GFS requires the CMA-GFS tangent linear model and its adjoint model, and the forecast quantity is the horizontal wind component in the CMA-GFS model forecast variables. ), disturbance potential temperature ( ) and disturbed dimensionless pressure ( The tangent linear disturbance variables of ) are expressed as follows: ; S45: For the adjoint gradient vector Implementation The transformation yields Eulerian space vectors; S46: Iterate through steps S41 to S45 to obtain the fastest growing Euler vector as the iteration termination condition, and the fastest growing Euler vector is the singular vector SVs. S5: Generate the sensitive region based on the linear combination of the singular vectors SVs; The formula for calculating the sensitive area is: , In the formula, For the calculated first A singular value, For the corresponding number A singular vector; S6: Draw a display diagram of the sensitive area.

2. The method according to claim 1, characterized in that, The singular vector SVs perturbation is a set of orthogonal perturbations that grow the fastest within a certain time interval based on the total energy transformation operator E in the tangent linear mode.

3. The method according to claim 2, characterized in that, The formula adopted by the total energy transformation operator E-transformation is: ， In the formula, The accompanying mode of the CMA-GFS tangent linear mode. For matrix operators The i-th singular value, , For total energy transformation operator; It is the transpose of P, and P is the target region projection operator.

4. The method according to claim 2, characterized in that, The total energy weighting modulus The calculation formula is as follows: ， In the formula, the sum of the first two terms represents the perturbation kinetic energy, and the sum of the last two terms represents the perturbation potential energy. The third and fourth terms represent the perturbation potential temperature in the potential energy, respectively. and dimensionless atmospheric pressure disturbance The contribution of the quantity; In the above formula ,in To track coordinates for terrain, and These represent the longitude and latitude of the model's spherical coordinates, respectively. The specific heat at constant pressure of dry air; and These represent the reference temperature, reference potential temperature, reference dimensionless pressure, and reference density, respectively, and their calculation formulas are as follows: , , , The gas constant for dry air. For standard atmosphere, For reference temperature, This is the acceleration due to gravity.

5. The method according to claim 1, wherein the sensitive area display graph is a percentage graph.

6. A device for determining observation sensitive areas with high accuracy in weather forecasting, comprising: The determination module is used to determine the target area range and date; The acquisition module is used to acquire values ​​from the CMA-GFS 4DVAR business analysis field; The configuration module is used to interpolate the numerical values ​​of the business analysis field to the corresponding resolution for calculating singular vectors (SVs). The calculation module is used to calculate and generate the given value e. The steps performed by the calculation module include: The first step is to convert the Eulerian vectors into their corresponding vectors using the Lanczos algorithm. Perform total energy mode transformation operator The reverse Transform into initial perturbation vector ; The second step involves performing a parallel CMA-GFS tangent mode backward integration (time t) on the initial perturbation vector x to obtain the evolved perturbation vector. ; The third step is to... Perform total energy transformation operator The transformation yields the accompanying mode used as input to the CMA-GFS. input perturbation vector ; The fourth step is to process the accompanying perturbation vector. Using the parallel CMA-GFS adjoint mode Integrating forward by time interval t yields the adjoint gradient vector at the initial time. ; Among them, the CMA-GFS tangent linear model and its adjoint model required for CMA-GFS singular vector calculation have forecast quantities that are the horizontal wind components in the CMA-GFS model forecast variables. ), disturbance potential temperature ( ) and disturbed dimensionless pressure ( They are represented as follows: ; Fifth step, for the adjoint gradient vector Implementation The transformation yields Eulerian space vectors; Step 6: Iterate through steps 1 to 5, and use the fastest growing Euler vector as the termination condition for the iteration. The fastest growing Euler vector is taken as the singular vector SVs. The generation module generates sensitive regions based on a linear combination of the singular vectors SVs; The drawing module is used to draw a display diagram of the sensitive area.

7. A system for determining observation sensitive areas with high accuracy in weather forecasting, the system comprising: One or more processors; Storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement a method for determining observation sensitive areas with high accuracy in weather forecasting as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements a method for determining observation sensitive areas with high accuracy in weather forecasting as described in any one of claims 1-5.

Images