Data fusion method based on sentry, wind cloud satellite wind product and ERA5

Abstract

The invention provides a data fusion method based on a sentinel, a wind cloud satellite wind product and an ERA5, and belongs to the technical field of meteorological data processing and assimilation, and the method comprises the steps: obtaining a sentinel satellite offshore wind field data product, a wind cloud satellite offshore wind field data product and ERA5 reanalysis data of a to-be-observed region; preprocessing the sentinel satellite offshore wind field data product and the wind cloud satellite offshore wind field data product to obtain sentinel preprocessed offshore wind field data and wind cloud preprocessed offshore wind field data; constructing an analysis field by adopting the sentry preprocessed offshore wind field data, the wind cloud preprocessed offshore wind field data and the ERA5 reanalysis data to obtain a to-be-solved analysis field; and performing iterative optimization on the analysis field to be solved based on a three-dimensional assimilation method to obtain a fused wind field based on the sentinel satellite, the wind cloud satellite and the ERA5 data. According to the method, the sentry satellite, the wind cloud satellite and the ERA5 reanalysis data can be effectively fused, so that an analysis field with higher consistency is generated.

Description

Technical Field

[0001] The present invention relates to the technical field of meteorological data processing and assimilation, and in particular to a method, system, terminal and medium for fusing wind products of Sentinel and Fengyun satellites and ERA5 data. Background Art

[0002] In the early planning and construction of offshore wind farms, it is crucial to obtain accurate and comprehensive wind resource data. However, traditional observation methods, such as offshore wind towers, have problems such as insufficient data coverage and poor representativeness, and cannot meet the needs of large-scale offshore wind farm planning. At the same time, although floating lidar can provide high-precision wind resource data, its high cost and limited detection time limit its widespread application.

[0003] In order to solve the above problems, in recent years, the method of using remote sensing satellite observation data to evaluate offshore wind resources has gradually attracted attention. However, at present, the spatiotemporal weight method is usually used in satellite data fusion, that is, the data of the fused matching grid data set to be fused is fused based on the spatiotemporal weight to obtain the sea surface wind field fusion product corresponding to the study area. However, in the implementation process, this method lacks a meteorological background field and has no meteorological background field as a reference. The fusion between different satellite data will lead to inaccurate fusion results due to the lack of a unified atmospheric physics basis. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention proposes a method, system, terminal and medium based on Sentinel, Fengyun satellite wind products and ERA5 data fusion to solve the technical problems existing in the prior art.

[0005] The technical solution adopted by the present invention is:

[0006] The first aspect of this embodiment provides a method for fusing data based on Sentinel, Fengyun satellite wind products and ERA5, including:

[0007] Obtain Sentinel satellite offshore wind data products, Fengyun satellite offshore wind data products and ERA5 reanalysis data for the area to be observed;

[0008] Preprocessing the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data;

[0009] The Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data are used to construct the analysis field to obtain the analysis field to be solved;

[0010] The analysis field to be solved is iteratively optimized based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data.

[0011] Furthermore, the analysis field is constructed by using the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to obtain the analysis field to be solved, including:

[0012] Performing spatial matching and alignment on the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to obtain a spatial fusion product;

[0013] Calculate the wind speed deviation between the satellite data and the ERA5 reanalysis data in the spatial fusion product;

[0014] Performing polynomial interpolation processing on the wind speed deviation to obtain an interpolated wind speed deviation;

[0015] The hourly satellite wind speed data were calculated using the interpolated wind speed deviation and the wind speed values ​​of the ERA5 reanalysis data;

[0016] The constructed observation operator is used to correlate the hourly satellite wind speed data aligned by time and space matching with the ERA5 reanalysis data to obtain the analysis field to be solved.

[0017] Furthermore, the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data are spatially matched and aligned to obtain a spatial fusion product, including:

[0018] Based on the optimal interpolation model, the ERA5 reanalysis data is used as the background field, and the Sentinel pre-processed offshore wind field data and the Fengyun pre-processed offshore wind field data are spatially differencing fused to obtain a spatial fusion product. The ERA5 reanalysis data is the ERA5 reanalysis data with the most recent detection time of the Sentinel satellite and the Fengyun satellite.

[0019] Furthermore, the optimal interpolation model is expressed as:

[0020] x a =x b +K(y-Hx b )

[0021] Among them, x a is the analytical field after interpolation, x b is the background field, K is the gain matrix, y is the satellite wind speed observation data, and H is the observation operator;

[0022] Among them, the gain matrix:

[0023] K=BH T(HBH T +R) -1

[0024] Among them, B is the ERA5 data error covariance matrix, R is the satellite data observation error covariance matrix, H is the observation operator, and H T is the inverse matrix of the observation operator.

[0025] Furthermore, the three-dimensional assimilation method is used to iteratively optimize the analysis field to be solved, and obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data, including:

[0026] Construct a 3D-Var cost function that includes the analysis field to be solved, the ERA5 background field, the Sentinel and Fengyun satellite observation data, the background error covariance matrix, and the observation error covariance matrix;

[0027] Performing a derivative operation on the 3D-Var cost function to obtain a gradient of the 3D-Var cost function;

[0028] The conjugate gradient method is used to iteratively solve the gradient of the 3D-Var cost function, and the analysis field to be solved is updated. When the gradient of the 3D-Var cost function is within a preset range, a fused wind field based on the Sentinel satellite, Fengyun satellite and ERA5 data is obtained.

[0029] Furthermore, the 3D-Var cost function is expressed as:

[0030]

[0031] Where x is the analytical field to be solved; x b is the background field; o is the observed data; B is the background error covariance matrix; R is the observation error covariance matrix, and J(x) is the cost function value.

[0032] Furthermore, the gradient of the 3D-Var cost function is expressed as:

[0033]

[0034] in, Represents the gradient value of the 3D-Var cost function.

[0035] A second aspect of this embodiment provides a data fusion system based on Sentinel, Fengyun satellite wind products and ERA5, the system comprising:

[0036] The data acquisition module acquires the offshore wind data products of the Sentinel satellites, the offshore wind data products of the Fengyun satellites and the ERA5 reanalysis data of the area to be observed;

[0037] A data preprocessing module preprocesses the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data;

[0038] An analysis field construction module uses the Sentinel pre-processed offshore wind field data, Fengyun pre-processed offshore wind field data and ERA5 reanalysis data to construct an analysis field to obtain an analysis field to be solved;

[0039] The analysis field solution module iteratively optimizes the analysis field to be solved based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data.

[0040] A third aspect of this embodiment provides a terminal, including a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute any method described in the first aspect of this embodiment.

[0041] A fourth aspect of this embodiment provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the method as described in any one of the first aspects of this embodiment.

[0042] It can be seen from the above technical solution that the beneficial technical effects of the present invention are as follows:

[0043] 1. The present invention solves the problem of differences in space, time and frequency of observation data from different satellites. By effectively integrating Sentinel satellites, Fengyun satellites and ERA5 reanalysis data, the continuity in space and time is enhanced, the influence of observation errors is reduced, and a more consistent analysis field is generated.

[0044] 2. The present invention fuses the high spatial resolution data generated by Sentinel, Fengyun satellite data and ERA5 reanalysis data through the 3D-Var (three-dimensional variational assimilation) method. When locating offshore wind farm sites, it can more accurately identify areas with higher wind speeds and avoid misjudgment caused by low-resolution data, thereby optimizing the layout and design of wind farms. At the same time, it can provide support for the prediction of wind farm power generation capacity and help formulate more scientific and reasonable power generation planning and grid access strategies.

[0045] 3. The present invention can improve the accuracy of deep-sea wind resource assessment. The deep-sea wind resource assessment faces the lack of observation data and complex meteorological conditions. Sentinel satellite data and Fengyun satellite data can provide reliable observation wind speed under complex meteorological conditions such as cloudy and foggy weather. Combining Sentinel satellite, Fengyun satellite and ERA5 data can more comprehensively reflect the wind resource conditions in deep-sea areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the prior art description. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0047] Figure 1 This is a flow chart of an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of deviation interpolation according to an embodiment of the present invention;

[0049] Figure 3 A schematic diagram of a system according to an embodiment of the present invention;

[0050] Figure 4 Schematic diagram of a terminal according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0052] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0053] Example

[0054] This embodiment provides a method for fusing data based on Sentinel, Fengyun satellite wind products and ERA5. Figure 1 As shown;

[0055] The working principle of Example 1 is described in detail below:

[0056] S1: Obtain the Sentinel satellite offshore wind data products, Fengyun satellite offshore wind data products and ERA5 reanalysis data in the area to be observed;

[0057] In one embodiment, the Sentinel-1 satellite sea surface 10m wind speed inversion product, the Fengyun-3E satellite 10m wind speed inversion product and the ERA5 reanalysis 10m wind speed and wind direction hourly data of the data study sea area are obtained.

[0058] Among them, the Sentinel-1 satellite ocean wind field product (Ocean Wind Field, OCN), the OCN product is the inversion of wind field (including wind speed and direction) by the "geophysical model function"; the microwave scatterometer (Wind Radar) and microwave imager (Microwave Imager) carried by the Fengyun-3E satellite measure the microwave scattering and radiation characteristics of the ocean surface to invert the wind speed and direction at a height of 10 meters above the sea surface; ERA5 data is the latest generation of global reanalysis data sets released by the European Centre for Medium-Range Weather Forecasts (ECMWF), which provides hourly wind speed and wind direction variables from 1979 to the present, with a horizontal resolution of 25 kilometers.

[0059] S2: preprocessing the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data;

[0060] In one embodiment, the acquired satellite data is preprocessed, including outlier processing, missing value filling, data consistency correction, noise reduction processing, and final verification and output.

[0061] Threshold setting: set the wind speed range to 0-70m / s and remove abnormal values ​​outside the range.

[0062] Statistical-based outlier detection: Calculate the mean and standard deviation of the data;

[0063] The formula for calculating the mean of the data is expressed as:

[0064]

[0065] The formula for calculating the standard deviation of data is expressed as:

[0066]

[0067] Among them, V i represents wind speed, μ represents average wind speed, σ represents standard deviation, and N represents data statistics;

[0068] Data with a multiple of more than 3 standard deviations are considered abnormal, and data that meets the condition (μ+k*σ)≤V≤(μ+k*σ) are retained (usually k is 3).

[0069] S3: constructing an analysis field using the Sentinel pre-processed offshore wind field data, Fengyun pre-processed offshore wind field data and ERA5 reanalysis data to obtain an analysis field to be solved;

[0070] In one embodiment, the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data are used to construct an analysis field to obtain an analysis field to be solved, including:

[0071] Performing spatial matching and alignment on the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to obtain a spatial fusion product;

[0072] Calculate the wind speed deviation between the satellite data and the ERA5 reanalysis data in the spatial fusion product;

[0073] Performing polynomial interpolation processing on the wind speed deviation to obtain an interpolated wind speed deviation;

[0074] The hourly satellite wind speed data were calculated using the interpolated wind speed deviation and the wind speed values ​​of the ERA5 reanalysis data;

[0075] The constructed observation operator is used to correlate the hourly satellite wind speed data aligned by time and space matching with the ERA5 reanalysis data to obtain the analysis field to be solved.

[0076] Specifically, in one embodiment, since the spatial resolution of Sentinel-1 satellite data is 500m×500m, and the spatial resolution of Fengyun satellite and ERA5 data is 25km×25km, considering that the fusion of the three data is spatially fused, which is a multi-scale and multi-resolution data fusion problem, all data are adjusted to a unified spatial grid, and the OCN (Ocean) product data of Sentinel-1 and the GNOS-II sea surface wind speed data of Fengyun-3 satellite (FY-3E) are interpolated to the ERA5 grid by using the optimal interpolation method to retain finer spatial details.

[0077] In one embodiment, the spatial matching and alignment of the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to obtain a spatial fusion product includes:

[0078] Based on the detection time of the Sentinel satellites and the Fengyun satellites, the ERA5 reanalysis data closest to the detection time of the Sentinel satellites and the Fengyun satellites are obtained. If the time difference is greater than or equal to 30 minutes, the average field of the ERA5 reanalysis data at the two moments before and after is selected;

[0079] Based on the optimal interpolation model, the ERA5 reanalysis data is used as the background field, and the Sentinel pre-processed offshore wind field data and the Fengyun pre-processed offshore wind field data are spatially difference fused to obtain a spatial fusion product.

[0080] The optimal interpolation model is expressed as:

[0081] x a =x b +K(y-Hx b )

[0082] Among them, x a is the analytical field after interpolation, x b is the background field, K is the gain matrix, y is the satellite wind speed observation data, and H is the observation operator;

[0083] Among them, the gain matrix:

[0084] K=BH T (HBH T +R) -1

[0085] Among them, B is the ERA5 data error covariance matrix, R is the satellite data observation error covariance matrix, H is the observation operator, and H T is the inverse matrix of the observation operator.

[0086] Based on the optimal interpolation model, with ERA5 as the background field, the satellite data (Sentinel satellites, Fengyun satellites) are spatially interpolated and fused to form a spatial fusion product.

[0087] Calculate the wind speed deviation between the satellite data and the ERA5 reanalysis data in the spatial fusion product;

[0088] In one embodiment, considering the poor temporal resolution of satellite data, the deviation polynomial interpolation method is used for temporal interpolation. The wind speed deviation between satellite data and ERA5 data is calculated by using the spatial fusion product obtained in the previous step and ERA5 as the background field through the corresponding time and space points;

[0089] Performing polynomial interpolation processing on the wind speed deviation to obtain an interpolated wind speed deviation;

[0090] In one embodiment, polynomial interpolation is performed on the wind speed deviation to obtain an interpolated wind speed deviation, which is expressed as:

[0091]

[0092] Where B(t) represents the deviation between satellite fusion data and ERA5 reanalysis data at time t, y i represents the deviation between the satellite fusion data and the ERA5 reanalysis data at time i, t represents the time (UTC), t i and t j They represent time i and time j respectively, and n represents the total number of detection time points of satellite fusion data.

[0093] After completing the polynomial bias interpolation, the interpolated wind speed bias and the wind speed value of the ERA5 reanalysis data are used to perform back calculations to output hourly satellite data:

[0094] R(i)=Wind(i)+B(i)

[0095] Where R represents the hourly continuous fused wind speed data value, Wind represents the ERA5 wind speed value at time i, and B(i) represents the deviation value between the satellite fusion data at time t and the ERA5 reanalysis data at time i, as shown in Figure 2 shown.

[0096] The constructed observation operator is used to correlate the hourly satellite wind speed data aligned by time and space matching with the ERA5 reanalysis data to obtain the analysis field to be solved.

[0097] In one embodiment, the time points of the spatially fused single satellite data are aligned to unity (hourly time step), and the hourly series data are reanalyzed using ERA5 as the background field, which has a high temporal resolution and a good long-term trend. The wind speed field of ERA5 at time t is represented by xb(t), and the wind speed data of the spatiotemporally fused Sentinel and Fengyun satellites are used as observation data, recorded as yo(t). Subsequently, an observation operator H is constructed to associate the observation data with the background field data to obtain the analysis field to be solved.

[0098] S4: Iteratively optimize the analysis field to be solved based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data;

[0099] In one embodiment, the iterative optimization of the analysis field to be solved based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data includes:

[0100] Construct a 3D-Var cost function that includes the analysis field to be solved, the ERA5 background field, the Sentinel and Fengyun satellite observation data, the background error covariance matrix, and the observation error covariance matrix;

[0101] Performing a derivative operation on the 3D-Var cost function to obtain a gradient of the 3D-Var cost function;

[0102] The conjugate gradient method is used to iteratively solve the gradient of the 3D-Var cost function, and the analysis field to be solved is updated. When the gradient of the 3D-Var cost function is within a preset range (the preset value range is 10 -4 -10 -5 The fused wind field based on Sentinel, Fengyun and ERA5 data was obtained.

[0103] In one embodiment, the 3D-Var cost function is expressed as:

[0104]

[0105] Where x is the analytical field to be solved; x b is the background field; o is the observed data; B is the background error covariance matrix; R is the observation error covariance matrix, and J(x) is the cost function value.

[0106] In one embodiment, the gradient of the 3D-Var cost function is expressed as:

[0107]

[0108] in, Represents the gradient value of the 3D-Var cost function.

[0109] In one embodiment, a data fusion system based on Sentinel, Fengyun satellite wind products and ERA5 is provided, such as Figure 3 As shown, the system comprises:

[0110] The data acquisition module 301 acquires the offshore wind data products of the Sentinel satellite, the offshore wind data products of the Fengyun satellite and the ERA5 reanalysis data of the area to be observed;

[0111] The data preprocessing module 302 preprocesses the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data;

[0112] The analysis field construction module 303 uses the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to construct the analysis field to obtain the analysis field to be solved;

[0113] The analysis field solving module 304 iteratively optimizes the analysis field to be solved based on a three-dimensional assimilation method to obtain a fused wind field based on the Sentinel satellite, Fengyun satellite and ERA5 data.

[0114] like Figure 4 As shown, an embodiment of the present application provides a terminal, including a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, and the computer program includes program instructions, and the processor is configured to call the program instructions to execute part or all of the steps described in an embodiment of the present application, a method for fusion of data based on Sentinel, Fengyun satellite wind products and ERA5.

[0115] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program, and the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes part or all of the steps described in an embodiment of the present application, a method for fusion of data based on Sentinel, Fengyun satellite wind products and ERA5.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data, characterized in that: include: Obtain Sentinel satellite offshore wind data products, Fengyun satellite offshore wind data products and ERA5 reanalysis data for the area to be observed; Preprocessing the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data; The Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data are used to construct the analysis field to obtain the analysis field to be solved; The analysis field to be solved is iteratively optimized based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data.

2. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 1, characterized in that: The analysis field is constructed by using the Sentinel pre-processed offshore wind field data, Fengyun pre-processed offshore wind field data and ERA5 reanalysis data to obtain the analysis field to be solved, including: Performing spatial matching and alignment on the Sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data to obtain a spatial fusion product; Calculate the wind speed deviation between the satellite data and the ERA5 reanalysis data in the spatial fusion product; Performing polynomial interpolation processing on the wind speed deviation to obtain an interpolated wind speed deviation; The hourly satellite wind speed data were calculated using the interpolated wind speed deviation and the wind speed values ​​of the ERA5 reanalysis data; The constructed observation operator is used to correlate the hourly satellite wind speed data aligned by time and space matching with the ERA5 reanalysis data to obtain the analysis field to be solved.

3. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 2, characterized in that: The sentinel pre-processed offshore wind farm data, the Fengyun pre-processed offshore wind farm data and the ERA5 reanalysis data are spatially matched and aligned to obtain a spatial fusion product, including: Based on the optimal interpolation model, the ERA5 reanalysis data is used as the background field, and the Sentinel pre-processed offshore wind field data and the Fengyun pre-processed offshore wind field data are spatially differencing fused to obtain a spatial fusion product. The ERA5 reanalysis data is the ERA5 reanalysis data with the most recent detection time of the Sentinel satellite and the Fengyun satellite.

4. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 3, characterized in that: The optimal interpolation model is expressed as: x a =x b +K(y-Hx b ) Among them, x a is the analytical field after interpolation, x b is the background field, K is the gain matrix, y is the satellite wind speed observation data, and H is the observation operator; Among them, the gain matrix: K=BH T (HBH T +R) -1 Among them, B is the ERA5 data error covariance matrix, R is the satellite data observation error covariance matrix, H is the observation operator, and H T is the inverse matrix of the observation operator.

5. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 1, characterized in that: The three-dimensional assimilation method is used to iteratively optimize the analysis field to be solved, and obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data, including: Construct a 3D-Var cost function that includes the analysis field to be solved, the ERA5 background field, the Sentinel and Fengyun satellite observation data, the background error covariance matrix, and the observation error covariance matrix; Performing a derivative operation on the 3D-Var cost function to obtain a gradient of the 3D-Var cost function; The conjugate gradient method is used to iteratively solve the gradient of the 3D-Var cost function, and the analysis field to be solved is updated. When the gradient of the 3D-Var cost function is within a preset range, a fused wind field based on the Sentinel satellite, Fengyun satellite and ERA5 data is obtained.

6. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 5, characterized in that: The 3D-Var cost function is expressed as: Where x is the analytical field to be solved; x b is the background field; o is the observed data; B is the background error covariance matrix; R is the observation error covariance matrix, and J(x) is the cost function value.

7. The method for fusion of Sentinel, Fengyun satellite wind products and ERA5 data according to claim 6, characterized in that: The gradient of the 3D-Var cost function is expressed as: ▽J(x)=B -1 (x-x b )+H T R -1 (H(x)-y) Among them, ▽J(x) represents the gradient value of the 3D-Var cost function.

8. A data fusion system based on Sentinel, Fengyun satellite wind products and ERA5, characterized in that: The system comprises: The data acquisition module acquires the offshore wind data products of the Sentinel satellites, the offshore wind data products of the Fengyun satellites and the ERA5 reanalysis data of the area to be observed; A data preprocessing module preprocesses the Sentinel satellite offshore wind farm data product and the Fengyun satellite offshore wind farm data product to obtain Sentinel preprocessed offshore wind farm data and Fengyun preprocessed offshore wind farm data; An analysis field construction module uses the Sentinel pre-processed offshore wind field data, Fengyun pre-processed offshore wind field data and ERA5 reanalysis data to construct an analysis field to obtain an analysis field to be solved; The analysis field solution module iteratively optimizes the analysis field to be solved based on the three-dimensional assimilation method to obtain a fused wind field based on Sentinel satellite, Fengyun satellite and ERA5 data.

9. A terminal, characterized in that: The method comprises a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 7.

Images