Method for evaluating influence of meteorological observation data on numerical forecasting and related equipment

Abstract

The invention discloses a method for evaluating the influence of meteorological observation data on numerical forecasting and related equipment, and relates to the technical field of meteorological data evaluation, and the method comprises the steps: obtaining observation data and meteorological grid point data; performing interpolation on the meteorological grid point data to obtain low-resolution data; inputting the low-resolution data at the initial moment into a CMA-GFS system to obtain a forecast result; respectively calculating a forecast error taking the background field as an initial value, a forecast error taking the analysis field as an initial value and difference data of the two forecast errors; calculating a background field gradient, an analysis field gradient and an observation data gradient of a model of the forecast error; and finally, obtaining the contribution of the observation data to forecast error reduction according to the gradient of the model of the forecast error to the observation data and the variation of the observation. According to the method, the contribution of any observation data to prediction error reduction at any position can be quantitatively evaluated, and the sensitivity of prediction to observation is obtained.

Description

Technical Field

[0001] The present application relates to the technical field of meteorological data evaluation, and in particular to a method for evaluating the impact of meteorological observation data on numerical forecasting and related equipment. Background Art

[0002] Based on the sensitivity analysis of the forecast relative to the observation (FSO), it uses the basic operators in the 4D variational data assimilation system to quantify the sensitivity of a certain observation to the analysis or forecast error. The biggest advantage of this method is that all observations used in the assimilation system can be evaluated at the same time, so FSO will be a very good way to monitor the contribution of different types of observations. At present, many related studies based on their respective business models have been carried out, but we do not have a method in our business that can comprehensively evaluate all data at once. Summary of the invention

[0003] The purpose of this application is to provide a method and related equipment for evaluating the impact of meteorological observation data on numerical forecasts, which can be used to comprehensively evaluate meteorological data.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for evaluating the impact of meteorological observation data on numerical forecasts, the method for evaluating the impact of meteorological observation data on numerical forecasts comprising:

[0006] Acquiring observation data and meteorological grid data; the meteorological grid data includes: high-resolution background field data and high-resolution analysis field data;

[0007] Interpolating the meteorological grid data to obtain low-resolution data; the low-resolution data includes: low-resolution data at the initial moment and true value analysis field data; the low-resolution data at the initial moment includes: low-resolution background field data at the initial moment and low-resolution analysis field data at the initial moment; the true value analysis field data is low-resolution analysis field data at the forecast moment;

[0008] Inputting the low-resolution data at the initial moment into the CMA-GFS system to obtain a forecast result; the forecast result includes a first forecast field with the low-resolution background field data at the initial moment as an initial value and a second forecast field with the low-resolution analysis field data at the initial moment as an initial value;

[0009] Subtracting the first forecast field from the true value analysis field data to obtain a first forecast error;

[0010] Subtracting the second forecast field from the true value analysis field data to obtain a second forecast error;

[0011] Calculating a difference between the modulus of the first forecast error and the modulus of the second forecast error to obtain a change in the forecast error;

[0012] Using the modulus of the first prediction error, calculate the gradient of the modulus of the prediction error relative to the low-resolution background field data at an initial moment to obtain a first gradient;

[0013] Using the modulus of the second forecast error, calculating the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at an initial moment, to obtain a second gradient;

[0014] Calculate the gradient of the magnitude of the prediction error relative to the observed data according to the first gradient and the second gradient;

[0015] The contribution of the observed data to the reduction of the forecast error is calculated based on the gradient and the change of the observed data.

[0016] Optionally, obtain meteorological grid data, including:

[0017] Get the original observation data;

[0018] The original observation data are assimilated to obtain meteorological grid data.

[0019] Optionally, the original observation data includes:

[0020] Sounding data, aircraft reporting data, cloud wind data, ocean surface wind data, reflectometer wind data, ship reporting data, ground reporting data, GPS precipitable water data, occultation data, microwave thermometer data, hygrometer data, imager data and hyperspectral data.

[0021] Optionally, using the modulus of the first prediction error to calculate the gradient of the modulus of the prediction error relative to the low-resolution background field data at an initial moment to obtain the first gradient specifically includes:

[0022] Calculating the gradient of the modulus of the forecast error relative to the first forecast field using the modulus of the first forecast error to obtain a background field forecast gradient;

[0023] The gradient of the modulus of the prediction error relative to the low-resolution background field data at the initial moment is calculated according to the background field prediction gradient to obtain a first gradient.

[0024] Optionally, using the modulus of the second forecast error to calculate the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at an initial moment to obtain the second gradient specifically includes:

[0025] Calculating the gradient of the modulus of the forecast error relative to the second forecast field using the modulus of the second forecast error to obtain the analysis field forecast gradient;

[0026] The gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment is calculated according to the analysis field forecast gradient to obtain a second gradient.

[0027] Optionally, respectively calculating the gradient of the modulus of the prediction error relative to the observed data according to the first gradient and the second gradient includes:

[0028] Solve the linear equations using the Lanczos-CG algorithm: or The solution result is obtained; among them,

[0029] The solution result is used to calculate the gradient of the model of the forecast error relative to the observed data. The calculation formula is:

[0030] Optionally, the interval between the initial time shown and the forecast time is 24 hours.

[0031] In a second aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for evaluating the impact of meteorological observation data on numerical forecasts as described above.

[0032] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for evaluating the impact of meteorological observation data on numerical forecasts as described in any one of the above.

[0033] In a fourth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for evaluating the impact of meteorological observation data on numerical forecasts as described in any one of the above.

[0034] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0035] The present application provides an evaluation method and related equipment for the impact of meteorological observation data on numerical forecasting, the method comprising: acquiring observation data and meteorological grid data; the meteorological grid data comprising: high-resolution background field data and high-resolution analysis field data; interpolating the meteorological grid data to obtain low-resolution data; the low-resolution data comprising: low-resolution data at the initial moment and true value analysis field data; the low-resolution data at the initial moment comprising: low-resolution background field data at the initial moment and low-resolution analysis field data at the initial moment; the true value analysis field data is low-resolution analysis field data at the forecast moment; inputting the low-resolution data at the initial moment into a CMA-GFS system to obtain a forecast result; the forecast result comprises a first forecast field with the low-resolution background field data at the initial moment as an initial value and a low-resolution analysis field data at the initial moment. The second forecast field with field data as the initial value; using the first forecast field to subtract the true value analysis field data to obtain the first forecast error; using the second forecast field to subtract the true value analysis field data to obtain the second forecast error; calculating the difference between the modulus of the first forecast error and the modulus of the second forecast error to obtain the change of the forecast error; using the modulus of the first forecast error to calculate the gradient of the modulus of the forecast error relative to the low-resolution background field data at the initial moment to obtain the first gradient; using the modulus of the second forecast error to calculate the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment to obtain the second gradient; calculating the gradient of the modulus of the forecast error relative to the observed data based on the first gradient and the second gradient; calculating the contribution of the observed data to the reduction of the forecast error based on the change of the gradient and the observed data. The present application can conduct a comprehensive quantitative assessment of the impact of meteorological data forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A schematic diagram of the basic principle of FSO provided in an embodiment of the present application;

[0038] Figure 2 This is an application environment diagram of a method for evaluating the impact of meteorological observation data on numerical forecasting in one embodiment of the present application;

[0039] Figure 3 A schematic diagram of a flow chart of a method for evaluating the impact of meteorological observation data on numerical forecasting provided in one embodiment of the present application;

[0040] Figure 4A schematic diagram of a data processing flow of a method for evaluating the impact of meteorological observation data on numerical forecasting provided in one embodiment of the present application;

[0041] Figure 5 A schematic diagram of the FSO evaluation results of the CMA-GFS4.0 assimilation application observation data in September 2023 provided in one embodiment of the present application;

[0042] Figure 6 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0044] Depend on Figure 1 It can be seen that changes in observations can cause changes in forecast errors, and FSO can quantitatively evaluate the reduction in forecast errors for each observation, thereby accurately evaluating the benefits of related observations. The biggest advantage of this method is that it can evaluate all observations used in the assimilation system at the same time. Figure 1 The definitions of the variables are shown in Table 1. Baker and Daley (Baker and Daley, 2000) first proposed using the adjoint of the analytical system to calculate the sensitivity relative to observations, while Langland and Baker (Langland and Baker, 2004) used the adjoint model to calculate the sensitivity of the forecast relative to observations, ultimately achieving a connection between observed changes and forecast errors.

[0045] Table 1 Variable definition table

[0046]

[0047] The purpose of this application is to provide a method and related equipment for evaluating the impact of meteorological observation data on numerical forecasts, which can be used to comprehensively evaluate meteorological data.

[0048] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0049] The method for evaluating the impact of meteorological observation data on numerical forecasting provided in the embodiment of the present application can be applied to Figure 2In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the meteorological grid data to be processed to the server 104. After receiving the meteorological grid data to be processed, the server 104 interpolates the meteorological grid data to obtain low-resolution data; the low-resolution data includes: low-resolution data at the initial moment and true value analysis field data; the low-resolution data at the initial moment includes: low-resolution background field data at the initial moment and low-resolution analysis field data at the initial moment; the true value analysis field data is low-resolution analysis field data at the forecast moment; the low-resolution data at the initial moment is input into the CMA-GFS system to obtain a forecast result; the forecast result includes a first forecast field with the low-resolution background field data at the initial moment as the initial value and a second forecast field with the low-resolution analysis field data at the initial moment as the initial value; using The first forecast field is subtracted from the true value analysis field data to obtain a first forecast error; the second forecast field is subtracted from the true value analysis field data to obtain a second forecast error; the difference between the modulus of the first forecast error and the modulus of the second forecast error is calculated to obtain a change in the forecast error; the modulus of the first forecast error is used to calculate the gradient of the modulus of the forecast error relative to the low-resolution background field data at the initial moment to obtain a first gradient; the modulus of the second forecast error is used to calculate the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment to obtain a second gradient; the gradient of the modulus of the forecast error relative to the observation data is calculated based on the first gradient and the second gradient; the contribution of the observation data to the reduction of the forecast error is calculated based on the gradient and the change in the observation data. The server 104 can feed back the obtained result to the terminal 102. In addition, in some embodiments, the evaluation method of the meteorological grid data can also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 can directly process the meteorological grid data to be processed, or the server 104 can obtain the meteorological grid data to be processed from the data storage system and process it.

[0050] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0051] In an exemplary embodiment, Figure 3 and Figure 4As shown, a method for evaluating the impact of meteorological observation data on numerical forecasting is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 2 The server 104 in the example is used as an example to illustrate the method, which includes the following steps S1 to S10. Among them:

[0052] S1. Obtain observation data and meteorological grid data; the meteorological grid data includes: high-resolution background field data and high-resolution analysis field data. In this embodiment, the grid data obtained by the forecasting system is interpolated.

[0053] The meteorological grid data include: sounding data, aircraft reporting data, cloud guide wind data, ocean surface wind data, reflectometer wind data, ship reporting data, ground reporting data, GPS precipitable water data, occultation data, microwave thermometer data, hygrometer data, imager data and hyperspectral data.

[0054] At present, the observation data used in the CMA-GFS4.0 assimilation system independently developed by the China Earth System Numerical Prediction Center include: sounding (temperature, humidity, U and V wind), aircraft report (temperature, U and V wind), cloud guide wind (U and V wind), ocean surface wind (U and V wind), reflectometer wind (wind speed), ship report, ground report, GPS precipitable water, occultation (FY3C, FY3D, FY3E, GRAC, GRAD, TSAR, MP01, MP03, KOMS, TDMX, C2E1, C2E2, C2E3, C2E4, C2E5, C2E6, SE6A, PAZ1, GEOP, SPIR, YX2A, YX2B, YX2C, YX2D, YX2E, Y X2F, TM00, TM02, TM03, TM04, TM05 and TM06), microwave thermometers (NOAA15, NOAA18, NOAA19, NPP1, NOAA20, FY3D and FY3E) and hygrometers (FY3C, FY3D, FY3E, METOPB, METOPC, NOAA19, NPP1 and NOAA20), imagers (FY4A, FY2H, FY3D, GCOW and GOES16), hyperspectral (METOPB, METOPC, NOAA20, FY3D, FY4A and FY4A2), etc., with a single-time data usage of 1.5-2 million copies and a daily usage of more than 6 million copies. FSO can evaluate the application effect of all the above-mentioned observation data. Figure 2The FSO evaluation results of various observation data are given. Among them, aircraft reports have the greatest effect on reducing the global model forecast error, with U+V+T contributing up to 30%, followed by cloud guide wind, with U+V contributing more than 13.7%, and occultation data also make a large positive contribution, accounting for 8.7%.

[0055] Among them, U and V winds represent the horizontal east-west winds and north-south winds. In the above text, the ones starting with FY are my country's Fengyun satellites. If the Fengyun satellite is followed by an even number, it means a geostationary satellite. If it is an odd number, it is a polar-orbiting satellite. ABCD, etc., represent the first and second stars of this type of satellite. For example, FY4A represents the first Fengyun geostationary satellite. The ones starting with TM are Tianmu satellites, and the ones starting with YX are Yunyao satellites. NOAA** are polar-orbiting satellites launched by the United States at different times. METOP is a European polar-orbiting satellite. Many satellites in the occultation data are small satellites, including those from Europe, Asia, and America.

[0056] The step of obtaining meteorological grid data specifically includes: obtaining original observation data; and assimilating the original observation data to obtain meteorological grid data.

[0057] Specifically, in this embodiment, the high-resolution grapes_input obtained by assimilation into the entry model is the meteorological grid data mentioned above.

[0058] S2. Interpolate the meteorological grid data to obtain low-resolution data; the low-resolution data includes: low-resolution data at the initial moment and true value analysis field data; the low-resolution data at the initial moment includes: low-resolution background field data at the initial moment and low-resolution analysis field data at the initial moment; the true value analysis field data is low-resolution analysis field data at the forecast moment.

[0059] Interpolate the meteorological grid data to obtain low-resolution grapes_input, including the low-resolution background field x at time t0 bl and analysis field x al , t i The low-resolution analysis field x at the moment as the true value t al In this embodiment, time t0 is taken as the initial time, t i The time is used as the forecast time. The interpolation method is to perform three-dimensional linear interpolation calculation by calling an independent interpolation module.

[0060] S3. Input the low-resolution data at the initial moment into the CMA-GFS system to obtain a forecast result; the forecast result includes a first forecast field with the low-resolution background field data at the initial moment as the initial value and a second forecast field with the low-resolution analysis field data at the initial moment as the initial value.

[0061] Read in the x obtained from the first step bl and x al , integrating the low-resolution forecast model and the tangent linear model, we get two sets of t0 to t i The low-resolution model trajectory at time t (i.e., the background field forecast result) and i Low-resolution forecast field x at time fbl and x fal According to the test results, the current FSO interpolation system can only select the resolution of the internal and external loops used for four-dimensional change assimilation, and other resolutions are invalid.

[0062] S4. Subtract the first forecast field from the true value analysis field data to obtain a first forecast error.

[0063] Specifically, in this embodiment, the forecast result using the background field as the initial value is subtracted from the true value analysis field data at the corresponding time to obtain the forecast error using the background field as the initial value.

[0064] S5. Subtract the second forecast field from the true value analysis field data to obtain a second forecast error.

[0065] In this embodiment, the forecast field using the analysis field as the initial value is subtracted from the true value analysis field data at the corresponding moment to obtain the forecast error using the analysis field as the initial value.

[0066] S6. Calculate the difference between the modulus of the first forecast error and the modulus of the second forecast error to obtain a change in the modulus of the forecast error.

[0067] The difference between the modulus of the first forecast error and the modulus of the second forecast error is calculated to obtain a change in the forecast error.

[0068] Calculate the gradient of the low-resolution forecast error modulus with respect to the forecast field and Calculate the norm of the forecast error and its changes δe=e a -e b , where C represents a diagonal matrix consisting of low-resolution energy modulus coefficients. The forecast field is calculated, that is, the low-resolution analysis field is used as the initial value and the numerical forecast model is used for integration to obtain the forecast field after 24 hours.

[0069] S7. Using the modulus of the first prediction error, calculate the gradient of the modulus of the prediction error relative to the low-resolution background field data at an initial moment to obtain a first gradient.

[0070] In this embodiment, the modulus of the first forecast error is first used to calculate the gradient of the modulus of the forecast error relative to the first forecast field to obtain the background field forecast gradient;

[0071] Then, the gradient of the modulus of the prediction error relative to the low-resolution background field data at the initial moment is calculated according to the background field prediction gradient to obtain a first gradient.

[0072] S8. Using the modulus of the second forecast error, calculate the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment to obtain a second gradient.

[0073] In this embodiment, the modulus of the second forecast error is first used to calculate the gradient of the modulus of the forecast error relative to the second forecast field to obtain the analysis field forecast gradient;

[0074] Then, the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment is calculated according to the analysis field forecast gradient to obtain a second gradient.

[0075] Integrate the low-resolution adjoint model and calculate the gradient of the low-resolution forecast error modulus relative to the model initial value and

[0076] S9. Calculate the gradient of the modulus of the prediction error relative to the observed data based on the first gradient and the second gradient.

[0077] Read in the high-resolution background field x at time t0 b , integrate the 6-hour high-resolution forecast model and calculate the observation increment d = H(M(x b ))-y.

[0078] Solve the linear equations using the Lanczos-CG algorithm: or Get the solution result.

[0079] The solution result is used to calculate the gradient of the model of the forecast error relative to the observed data, and the calculation formula is:

[0080] S10. Calculate the contribution of the observed data to the reduction of the forecast error according to the gradient and the change of the observed data.

[0081] Calculate the effect of different observations on the magnitude of the forecast error Here i represents different observation types, different observation locations, and different observation variables. The observations here refer to the observations used by the CMA-GFS4.0 assimilation system, which are mentioned in the previous article. The impact of different observations on the model of the forecast error is the gradient of the model of the forecast error relative to each observation. Multiply by the observed change d i The observed change is equal to the observed value minus the observed value corresponding to the background field d i =yi -H i (x b ).

[0082] And compare δe and Output and

[0083] The calculation results are stored in the FSO_results directory in the form of files with the analysis time as the directory.

[0084] The web graphics production program developed can be used to draw various statistical charts for the files in the FSO result directory (such as Figure 5 shown).

[0085] In order to better illustrate the principles of the above steps, the following is an explanation:

[0086] The change in the magnitude of the forecast error δe and the change in the forecast field δx f The relationship between can be expressed as:

[0087]

[0088] Change of forecast field δx f The relationship between the change of the analytical field δx is:

[0089]

[0090] The relationship between the change in the analytical field δx and the change in the observation δy is:

[0091] δx=Kδy

[0092] The modulus e of the forecast error relative to the forecast field x f Gradient

[0093]

[0094] The gradient of the magnitude e of the prediction error with respect to the analysis field x

[0095]

[0096] When the disturbance approaches infinitesimal, the increment of the model of the prediction error δe is:

[0097]

[0098] When the disturbance does not approach infinitesimal, perform Taylor series expansion on δe

[0099]

[0100] First-order approximation:

[0101] Second order approximation:

[0102] Third order approximation:

[0103]

[0104] Errico (Errico, 2007) shortened the above formula to:

[0105]

[0106] According to the Lian rule, the gradient of the magnitude of the forecast error with respect to the observation is

[0107]

[0108] Where x = x b +K(yH(M(x b )).

[0109] Because K=U(I+U T M T H T R -1 HMU) -1 U T M T H T R -1 ,so

[0110]

[0111] In four-dimensional variational assimilation, the gradient of the objective functional is:

[0112]

[0113] Four-dimensional variational assimilation solves the equations iteratively using a minimization algorithm That is:

[0114]

[0115] Replace the right-hand side of the above equation with We can get a linear system of equations

[0116]

[0117] Using the existing minimization algorithm to solve

[0118]

[0119] After that, we can use the existing R, H, and M operators to calculate

[0120]

[0121] This embodiment converts the FSO text results into HDF5 data format through a program written in Python. Through the configuration file, it can be directly drawn on the web page. It is convenient for scientific researchers to use. It should be emphasized that the configuration file must be rewritten correctly according to the 4.0 observation data so that the newly added data can be correctly and completely displayed in the post-processing system. If the configuration file is incorrect, the FSO results drawn through the web page will have errors or missing items.

[0122] Based on this method, a one-time comprehensive evaluation of the application effect of global model observation data can be carried out. On the one hand, it can objectively give a scientific quantitative evaluation of each type, each layer and each data; on the other hand, it can give possible problems in the application of data, which is of great help to the improvement of my country's numerical forecast model assimilation technology. That is, the results are always evaluated, and the value corresponding to the first small i is the result of each observation. Based on these results, we can see the contribution of any observation at any point. If it is a positive contribution, it means that assimilation is effective. If it is a negative contribution, it means that there is a problem with assimilation and further improvement is needed.

[0123] This embodiment started a series of experiments based on the global model four-dimensional identification system, and finally preliminarily established an FSO system based on the CMA-GFS3.2 system in 2022. An FSO experiment based on the CMA-GFS3.3 system was conducted. In late May 2023, CMA-GFS4.0 was put into service, adding 11 types of observation data. The system was migrated from the original PI high-performance computing needs to the new high-performance computer, and the FSO system was fully updated from the operating environment, data interface, product processing, web page display, etc., forming the current CMA-GFS4.0 observation data FSO evaluation system. The current system has solved the defects of the original system's inability to operate across months (the script program has been modified and a new time function definition has been added, so that the original time definition has been increased from only one to two, with today and tomorrow. When crossing months, the two times can express the last day of the previous month and the first day of the next month, but the month definitions of these two times are different. The original definition was one, resulting in no corresponding file when one day is added to the last day of the month), unreasonable parallel node configuration (the parallel configuration has been modified according to the hardware conditions of the new machine. The original machine has 32 CPUs per node, but now it has 64. To configure for parallel operation with 64 CPUs, it is set to use 32 nodes, each with 64 CPUs), and unstable assimilation of low-resolution cycles (modify the convection parameterization and estimation correction parameter settings, turn off these two operating parameters, which were originally turned on and set to 0 or TRUE, but are now turned off and set to 1 or FALSE). The parallel output results have been synthesized, the program of the ground observation data conversion part of the mapping system has been modified, and the inaccurate application of some data in the assimilation system has been discovered and modified, ensuring that the new system runs stably in the new operating environment and the evaluation is accurate and reasonable.

[0124] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 6As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for evaluating the impact of meteorological observation data on numerical forecasting is implemented.

[0125] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0126] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0127] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0128] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0129] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0130] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0131] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0132] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for evaluating the impact of meteorological observation data on numerical forecasting, characterized in that: The evaluation method of the impact of meteorological observation data on numerical forecasting includes: Acquiring observation data and meteorological grid data; the meteorological grid data includes: high-resolution background field data and high-resolution analysis field data; Interpolating the meteorological grid data to obtain low-resolution data; the low-resolution data includes: low-resolution data at the initial moment and true value analysis field data; the low-resolution data at the initial moment includes: low-resolution background field data at the initial moment and low-resolution analysis field data at the initial moment; the true value analysis field data is low-resolution analysis field data at the forecast moment; Inputting the low-resolution data at the initial moment into the CMA-GFS system to obtain a forecast result; the forecast result includes a first forecast field with the low-resolution background field data at the initial moment as an initial value and a second forecast field with the low-resolution analysis field data at the initial moment as an initial value; Subtracting the first forecast field from the true value analysis field data to obtain a first forecast error; Subtracting the second forecast field from the true value analysis field data to obtain a second forecast error; Calculating a difference between the modulus of the first forecast error and the modulus of the second forecast error to obtain a change in the forecast error; Using the modulus of the first prediction error, calculate the gradient of the modulus of the prediction error relative to the low-resolution background field data at an initial moment to obtain a first gradient; Using the modulus of the second forecast error, calculating the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at an initial moment, to obtain a second gradient; Calculate the gradient of the magnitude of the prediction error relative to the observed data according to the first gradient and the second gradient; The contribution of the observed data to the reduction of the forecast error is calculated based on the gradient and the change of the observed data.

2. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 1, characterized in that: Obtain meteorological grid data, including: Get the original observation data; The original observation data are assimilated to obtain meteorological grid data.

3. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 2, characterized in that: The original observation data include: Sounding data, aircraft reporting data, cloud wind data, ocean surface wind data, reflectometer wind data, ship reporting data, ground reporting data, GPS precipitable water data, occultation data, microwave thermometer data, hygrometer data, imager data and hyperspectral data.

4. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 1, characterized in that: Calculating the gradient of the modulus of the prediction error relative to the low-resolution background field data at the initial moment using the modulus of the first prediction error to obtain a first gradient specifically includes: Calculating the gradient of the modulus of the forecast error relative to the first forecast field using the modulus of the first forecast error to obtain a background field forecast gradient; The gradient of the modulus of the prediction error relative to the low-resolution background field data at the initial moment is calculated according to the background field prediction gradient to obtain a first gradient.

5. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 1, characterized in that: Calculating the gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment by using the modulus of the second forecast error to obtain the second gradient specifically includes: Calculating the gradient of the modulus of the forecast error relative to the second forecast field using the modulus of the second forecast error to obtain the analysis field forecast gradient; The gradient of the modulus of the forecast error relative to the low-resolution analysis field data at the initial moment is calculated according to the analysis field forecast gradient to obtain a second gradient.

6. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 1, characterized in that: Calculating the gradient of the modulus of the forecast error relative to the observed data according to the first gradient and the second gradient respectively, specifically includes: Solve the linear equations using the Lanczos-CG algorithm: or The solution result is obtained; among them, The solution result is used to calculate the gradient of the model of the forecast error relative to the observed data. The calculation formula is:

7. The method for evaluating the impact of meteorological observation data on numerical forecasting according to claim 1, characterized in that: The interval between the initial time shown and the forecast time is 24 hours.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for evaluating the impact of meteorological observation data on numerical forecasting according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for evaluating the impact of meteorological observation data on numerical forecasting described in any one of claims 1 to 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for evaluating the impact of meteorological observation data on numerical forecasting described in any one of claims 1 to 7 is implemented.

Images