Construction and inversion method of sea surface wind speed inversion model based on dual-frequency wind field radar

Abstract

The invention discloses a construction and inversion method of a sea surface wind speed inversion model based on a dual-frequency wind field radar, the sea surface wind speed inversion model comprises a sea surface wind speed inversion model and a geophysical model function, and scattering measurement data, microwave radiometer data and sea surface wind field data are acquired; taking latitude and longitude and time information of the scattering measurement data as a reference, and respectively carrying out quasi-synchronous matching on the microwave radiometer data and the sea surface wind field data with the scattering measurement data by adopting a space-time matching threshold method to obtain an active and passive data set; on the basis of sea surface radiation and scattering characteristics, a geophysical model function for wind direction inversion and a sea surface wind speed inversion model capable of inhibiting rainfall influence are constructed by utilizing the characteristic that different frequencies have different penetrating power and adopting an active and passive spectrum gradient ratio, and the sea surface wind speed and the sea surface wind direction of a high-wind-speed sea surface can be obtained at the same time; the technical problem that it is difficult to quantitatively measure the sea surface wind speed and the sea surface wind direction of the high-wind-speed sea surface wind field at the same time in the prior art is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of space microwave remote sensing, relates to an inversion method, and specifically relates to a construction and inversion method of a sea surface wind speed inversion model based on a dual-frequency wind field radar. Background Technique

[0002] Tropical cyclones can cause strong winds or even super strong winds on the sea surface, often accompanied by strong convective weather phenomena such as heavy precipitation, lightning, and hail, seriously endangering the lives and property of residents in coastal areas affected by tropical cyclones. It is one of the most destructive natural disasters on Earth, and the maximum wind speed of typhoons can reach 70 - 80 m / s. Accurate high-wind-speed sea surface wind field information is of great significance for meteorological forecasting, disaster prevention and mitigation, and research on ocean thermal dynamic processes and ocean phenomena.

[0003] Currently, microwave remote sensing is the main means for large-scale and continuous measurement of the sea surface wind field. Under the action of the sea surface wind, the sea surface roughness increases. The microwave radar actively emits electromagnetic waves and measures the electromagnetic wave signals backscattered by the sea surface to obtain sea surface wind field information. The microwave radiometer passively measures the sea surface microwave radiation signal and inversely obtains sea surface wind information based on the relationship between microwave radiation and the sea surface wind.

[0004] Existing spaceborne microwave radars mainly operate in the C band and Ku band. Due to the characteristic that the co-polarized sea surface microwave scattering signal tends to saturate as the wind speed increases, the operational measurement range of the sea surface wind by spaceborne microwave scatterometers is usually 3 m / s - 24 m / s, and it is impossible to achieve high-precision quantitative measurement of the high-wind-speed sea surface wind field. On the other hand, the sea surface microwave radiation at high wind speeds mainly comes from the contribution of foam radiation. Compared with wind waves, the radiation rate of the sea surface foam does not easily tend to saturate with the change of wind speed and is approximately linear. Therefore, the microwave radiometer has the potential to measure high wind speeds. However, the lowest frequency of existing multi-frequency microwave radiometers usually operates in the C band and X band, and its electromagnetic wave wavelength is short, which is easily affected by rainfall. Currently, the maximum measurement ability of spaceborne microwave radiometers for operational sea surface wind is usually lower than 30 m / s.

[0005] In summary, the existing technology has the following problems:

[0006] (1) The existing sea surface wind field inversion algorithms based on the measurement data of spaceborne microwave scatterometers and radiometers are restricted by the physical characteristics of the sea surface, models, and inversion algorithms, and it is difficult to obtain high-wind-speed sea surface wind field information.

[0007] (2) The existing sea surface wind inversion algorithms based on the measurement data of spaceborne microwave scatterometers and microwave radiometers are usually independent of each other. The recently developed active-passive high-wind-speed joint inversion algorithm only targets wind speed inversion and it is difficult to obtain both wind speed and wind direction information at high wind speeds simultaneously.

[0008] Therefore, there is an urgent need to develop a high-wind-speed active-passive joint inversion method based on existing spaceborne microwave active and passive measurement data, and to provide sea surface wind speed and wind direction information at high wind speeds simultaneously. Summary of the Invention

[0009] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a method for constructing and inverting a sea surface wind speed inversion model based on a dual-frequency wind field radar, so as to solve the technical problem that it is difficult to simultaneously quantitatively measure the sea surface wind speed and sea surface wind direction of a high-wind-speed sea surface wind field in the existing technology.

[0010] To solve the above technical problems, the present invention is implemented by adopting the following technical solutions:

[0011] A method for constructing a sea surface wind field inversion model based on a dual-frequency wind field radar, the sea surface wind field inversion model includes a sea surface wind speed inversion model and a geophysical model function, and specifically includes the following steps:

[0012] Step 1, obtain scattering measurement data, microwave radiometer data, and sea surface wind field data; based on the longitude, latitude, and time information of the scattering measurement data, use the spatio-temporal matching threshold method to quasi-synchronously match the microwave radiometer data and the sea surface wind field data with the scattering measurement data respectively, and obtain the active-passive data set;

[0013] The scattering measurement data includes gridded sea surface backscattering coefficient data in the C band and its corresponding scattering measurement azimuth angle, and sea surface backscattering coefficient data in the Ku band and its corresponding scattering measurement azimuth angle;

[0014] Each grid includes multiple sea surface backscattering coefficient data in the C band and its corresponding scattering measurement azimuth angle, and multiple sea surface backscattering coefficient data in the Ku band and its corresponding scattering measurement azimuth angle;

[0015] The microwave radiometer data includes sea surface brightness temperature data in the C band, sea surface brightness temperature data in the X band, and sea surface brightness temperature data in the Ka band;

[0016] The sea surface wind field data includes sea surface wind speed and sea surface wind direction;

[0017] Step 2, construct the relationship between the scattering measurement data and the sea surface wind field data according to the active-passive data set, that is, obtain the geophysical model function σ 0，gm f ;

[0018]

[0019] Where:

[0020] Ws represents the sea surface wind speed, m / s;

[0021] Indicates the relative wind direction, which is the angle between the sea surface wind direction and the scattering measurement azimuth angle, °;

[0022] M represents the ratio of the active and passive spectral gradients;

[0023] σ 0,Cmeas Represents the sea surface backscattering coefficient data in the C band;

[0024] σ 0,Kumeas Represents the sea surface backscattering coefficient data in the Ku band;

[0025] TB C Represents the sea surface brightness temperature data in the C band, K;

[0026] TB Ka Represents the sea surface brightness temperature data in the Ka band, K;

[0027] a 0 a 1 and a 2 Are the coefficients of the geophysical model function σ 0 ;

[0028] Step 3: Use the maximum likelihood estimation method to search for the wind speed and wind direction that minimize the cost function J in the two-dimensional space of wind speed - wind direction as the ambiguous solution of the current grid, obtaining multiple ambiguous solutions corresponding to all grids. Use the circular median filtering method to find the optimal solutions for wind speed and wind direction among the multiple ambiguous solutions, that is, obtain the initial values of the sea surface wind direction and sea surface wind speed inversion;

[0029]

[0030] Where:

[0031] σ 0i,Cmeas Represents the sea surface backscattering coefficient data of the i-th C band in the current grid;

[0032] σ 0j,Kumeas Represents the sea surface backscattering coefficient data of the j-th Ku band in the current grid;

[0033] σ 0i,Cgmf Represents the geophysical mode value of the geophysical model function of the i-th C band in the current grid;

[0034] σ 0j,Kugmf Represents the geophysical mode value of the geophysical model function of the j-th Ku band in the current grid;

[0035] Var(σ 0j,Kumeas ) Represents the variance of the sea surface backscattering coefficient data of the i-th C band in the current grid;

[0036] Var(σ 0j,Kumeas) represents the variance of the backscattering coefficient data of the j-th Ku-band sea surface in the current grid;

[0037] i represents the serial number of the C-band falling into the current grid;

[0038] j represents the serial number of the Ku-band falling into the current grid;

[0039] Step 4: Using the microwave radiometer data in the active-passive dataset obtained in Step 1 and the initial value of sea surface wind speed inversion obtained in Step 3 as inputs, and the sea surface wind speed in the active-passive dataset obtained in Step 1 as the output, train the deep learning network model to obtain the sea surface wind speed inversion model.

[0040] The present invention further includes the following technical features:

[0041] The sea surface wind speed inversion model includes an input layer, a hidden layer, and an output layer;

[0042] The input layer is used to input parameters, and the parameters include microwave radiometer data and the initial value of sea surface wind speed inversion;

[0043] The hidden layer is used to map the corresponding relationship between the parameters and the sea surface wind speed data through an activation function;

[0044] The output layer is used to output the sea surface wind speed.

[0045] A method for jointly inversing the high-wind-speed sea surface wind field by active-passive combination based on a dual-frequency wind field radar specifically includes the following steps:

[0046] S1, collecting the sea surface active-passive dataset;

[0047] Collect scattering measurement data, microwave radiometer data, and sea surface wind field data; based on the longitude, latitude, and time information of the scattering measurement data, use the spatio-temporal matching threshold method to quasi-synchronously match the microwave radiometer data and the sea surface wind field data with the scattering measurement data respectively to obtain the sea surface active-passive dataset;

[0048] The scattering measurement data includes the gridded C-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth angle, and the Ku-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth angle;

[0049] Each grid includes multiple C-band sea surface backscattering coefficient data and their corresponding scattering measurement azimuth angles, and multiple Ku-band sea surface backscattering coefficient data and their corresponding scattering measurement azimuth angles;

[0050] The microwave radiometer data includes the C-band sea surface brightness temperature data, the X-band sea surface brightness temperature data, and the Ka-band sea surface brightness temperature data;

[0051] The sea surface wind field data includes sea surface wind speed and sea surface wind direction;

[0052] S2. Traverse the values in the wind speed - wind direction two - dimensional space, input the value into the geophysical model function obtained in step two of the construction method of the sea surface wind field inversion model based on the dual - frequency wind field radar, obtain the geophysical model value, and input it together with the sea surface passive - active data set obtained in S1 into the cost function J to obtain the initial values of the inverted sea surface wind direction and sea surface wind speed;

[0053] S3. Input the microwave radiometer data in the sea surface passive - active data set obtained in S1 and the initial value of the inverted sea surface wind speed obtained in S2 into the sea surface wind speed inversion model constructed by the construction method of the sea surface wind field inversion model based on the dual - frequency wind field radar for inversion to obtain the sea surface wind speed.

[0054] Compared with the prior art, the beneficial technical effects of the present invention are:

[0055] Based on the sea surface radiation and scattering characteristics, taking advantage of the different penetration capabilities of different frequencies, and adopting the passive - active spectral gradient ratio, the present invention constructs a geophysical model function for typhoon wind direction inversion and constructs a sea surface wind speed inversion model that can suppress the influence of rainfall. It can simultaneously obtain the sea surface wind speed and sea surface wind direction of the high - wind - speed sea surface, and solves the technical problem in the prior art that it is difficult to quantitatively measure the sea surface wind speed and sea surface wind direction of the high - wind - speed sea surface wind field at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is the flowchart of the method of the present invention;

[0057] Figure 2 is the structural schematic diagram of the sea surface wind speed inversion model in the present invention.

[0058] The following further elaborates on the specific content of the present invention in conjunction with embodiments. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] It should be noted that all components in the present invention, without special instructions, are components known in the art.

[0060] The following gives specific embodiments of the present invention. It should be noted that the present invention is not limited to the following specific embodiments, and all equivalent transformations made on the basis of the technical solutions of the present application fall within the protection scope of the present invention.

[0061] The present invention provides a construction method of a sea surface wind field inversion model based on a dual - frequency wind field radar. The sea surface wind field inversion model includes a sea surface wind speed inversion model and a geophysical model function, and specifically includes the following steps:

[0062] Step 1: Obtain scattering measurement data, microwave radiometer data, and sea surface wind field data; based on the longitude, latitude, and time information of the scattering measurement data, use the spatio-temporal matching threshold method to quasi-synchronously match the microwave radiometer data and the sea surface wind field data with the scattering measurement data respectively to obtain the active-passive data set;

[0063] The scattering measurement data includes gridded sea surface backscattering coefficient data in the C band and its corresponding scattering measurement azimuth angles, and sea surface backscattering coefficient data in the Ku band and its corresponding scattering measurement azimuth angles;

[0064] Each grid includes multiple sea surface backscattering coefficient data in the C band and their corresponding scattering measurement azimuth angles, and multiple sea surface backscattering coefficient data in the Ku band and their corresponding scattering measurement azimuth angles;

[0065] The microwave radiometer data includes sea surface brightness temperature data in the C band, sea surface brightness temperature data in the X band, and sea surface brightness temperature data in the Ka band;

[0066] The sea surface wind field data includes sea surface wind speed and sea surface wind direction;

[0067] Step 2: Establish the relationship between the scattering measurement data and the sea surface wind field data according to the active-passive data set, that is, obtain the geophysical model function σ 0，gm f ;

[0068]

[0069] Among them:

[0070] Ws represents the sea surface wind speed, m / s;

[0071] represents the relative wind direction, which is the angle between the sea surface wind direction and the scattering measurement azimuth angle, °;

[0072] M represents the active-passive spectral gradient ratio;

[0073] σ 0,Cmeas represents the sea surface backscattering coefficient data in the C band;

[0074] σ 0,Kumeas represents the sea surface backscattering coefficient data in the Ku band;

[0075] TB C represents the sea surface brightness temperature data in the C band, K;

[0076] TB Ka represents the sea surface brightness temperature data in the Ka band, K;

[0077] a 0 ,a 1 and a2 is the coefficient of the geophysical model function σ 0 ;

[0078] Step 3: Use the maximum likelihood estimation method to search in the two-dimensional space of wind speed - wind direction for the wind speed and wind direction that minimize the cost function J as the fuzzy solution of the current grid, obtaining multiple fuzzy solutions corresponding to all grids. Then, use the circular median filtering method to find the optimal solutions for wind speed and wind direction among the multiple fuzzy solutions, that is, obtain the initial values of the retrieved sea surface wind direction and sea surface wind speed;

[0079]

[0080] Where:

[0081] σ 0i,Cmeas represents the sea surface backscattering coefficient data of the i-th C-band in the current grid;

[0082] σ 0j,Kumeas represents the sea surface backscattering coefficient data of the j-th Ku-band in the current grid;

[0083] σ 0i,Cgmf represents the geophysical mode value of the geophysical model function of the i-th C-band in the current grid;

[0084] σ 0j,Kugmf represents the geophysical mode value of the geophysical model function of the j-th Ku-band in the current grid;

[0085] Var(σ 0j,Kumeas ) represents the variance of the sea surface backscattering coefficient data of the i-th C-band in the current grid;

[0086] Var(σ 0j,Kumeas ) represents the variance of the sea surface backscattering coefficient data of the j-th Ku-band in the current grid;

[0087] i represents the serial number of the C-band falling into the current grid;

[0088] j represents the serial number of the Ku-band falling into the current grid;

[0089] Step 4: Use the microwave radiometer data in the active - passive dataset obtained in Step 1 and the initial value of the retrieved sea surface wind speed obtained in Step 3 as inputs, and the sea surface wind speed in the active - passive dataset obtained in Step 1 as the output to train the deep learning network model, obtaining the sea surface wind speed retrieval model.

[0090] In the above technical solution, based on the radiation and scattering characteristics of the sea surface, taking advantage of the different penetration capabilities of different frequencies, and adopting the main passive spectral gradient ratio, a geophysical model function for typhoon wind direction inversion is constructed, and a sea surface wind speed inversion model that can suppress the influence of rainfall is constructed, which can simultaneously obtain the sea surface wind speed and sea surface wind direction of the high wind speed sea surface, solving the technical problem in the prior art that it is difficult to quantitatively measure the sea surface wind speed and sea surface wind direction of the high wind speed sea surface wind field at the same time.

[0091] In the above technical solution, a 0 , a 1 and a 2 are functions of the sea surface wind speed Ws and the main passive spectral gradient ratio M, and can be solved by fitting the relationship between the sea surface backscattering coefficient data and the sea surface wind speed Ws and the main passive spectral gradient ratio M.

[0092] The sea surface wind speed inversion model includes an input layer, a hidden layer and an output layer;

[0093] The input layer is used to input parameters, and the parameters include microwave radiometer data and the initial value of sea surface wind speed inversion;

[0094] The hidden layer is used to map the corresponding relationship between the parameters and the sea surface wind speed data through an activation function;

[0095] The output layer is used to output the sea surface wind speed.

[0096] A method for jointly inverting the main passive high wind speed sea surface wind field based on a dual-frequency wind field radar specifically includes the following steps:

[0097] S1, collect the main passive data set of the sea surface;

[0098] Collect scattering measurement data, microwave radiometer data and sea surface wind field data; based on the longitude, latitude and time information of the scattering measurement data, use the spatio-temporal matching threshold method to quasi-synchronously match the microwave radiometer data and the sea surface wind field data with the scattering measurement data respectively to obtain the main passive data set of the sea surface;

[0099] The scattering measurement data includes the gridded sea surface backscattering coefficient data in the C band and its corresponding scattering measurement azimuth angle and the sea surface backscattering coefficient data in the Ku band and its corresponding scattering measurement azimuth angle;

[0100] Each grid includes multiple sea surface backscattering coefficient data in the C band and their corresponding scattering measurement azimuth angles and multiple sea surface backscattering coefficient data in the Ku band and their corresponding scattering measurement azimuth angles;

[0101] The microwave radiometer data includes the sea surface brightness temperature data in the C band, the sea surface brightness temperature data in the X band and the sea surface brightness temperature data in the Ka band;

[0102] The sea surface wind field data includes sea surface wind speed and sea surface wind direction;

[0103] S2. Traverse the values in the wind speed - wind direction two - dimensional space, input the value into the geophysical model function obtained in step two of the construction method of the sea surface wind field inversion model based on the dual - frequency wind field radar, obtain the geophysical model value, and input it together with the sea surface active - passive data set obtained in S1 into the cost function J to obtain the initial values of the inverted sea surface wind direction and sea surface wind speed;

[0104] S3. Input the microwave radiometer data in the sea surface active - passive data set obtained in S1 and the initial value of the inverted sea surface wind speed obtained in S2 into the sea surface wind speed inversion model constructed by the construction method of the sea surface wind field inversion model based on the dual - frequency wind field radar for inversion to obtain the sea surface wind speed.

[0105] Comparative example:

[0106] Using the microwave radiometer data of OceanSat - 2B from January to October 2022, the wind field radar data of FY - 3E, and the sea surface high - speed inversion model of this application, the inverted sea surface wind speed Ws f and wind direction Wd f are obtained;

[0107] Adopting the wind speed Ws of the existing SMAP satellite wind speed data product smap and the wind direction Wd of the wind direction data product of the ERA5 model ERA5 to compare with the sea surface wind field inverted by the solution of this application, and calculate the statistical indexes of the inversion error:

[0108]

[0109]

[0110] Among them:

[0111] E bias_Ws represents the average deviation between the sea surface wind speed Ws of this application f and the wind speed Ws of the SMAP satellite wind speed data product smap ;

[0112] E bias_Wd represents the average deviation between the wind direction Wd of this application f and the wind direction Wd of the wind direction data product of the ERA5 model ERA5 ;

[0113] δ bias_Ws represents the standard deviation between the sea surface wind speed Ws of this application f and the wind speed Ws of the SMAP satellite wind speed data product smap ;

[0114] δ bias_Wd represents the wind direction Wd of the present application f and the wind direction Wd of the wind direction data product of the ERA5 model ERA5 standard deviation;

[0115] N represents the total number of statistical data.

[0116] It has been verified that the sea surface wind speed Ws of the present application of this solution f and the wind speed Ws of the SMAP satellite wind speed data product smap the average deviation of wind speed inversion is about 0.1 m / s. The sea surface wind speed Ws of this application f and the wind speed Ws of the SMAP satellite wind speed data product smap the standard deviation of wind speed inversion is about 2.2 m / s. The wind direction Wd of the present application f and the wind direction Wd of the wind direction data product of the ERA5 model ERA5 the average deviation of wind direction inversion is about 2.2°. The wind direction Wd of the present application f and the wind direction Wd of the wind direction data product of the ERA5 model ERA5 the standard deviation of wind direction inversion is about 25°. The maximum wind speed inversion ability that can be achieved is about 60 m / s, and the overall performance is better than the existing methods.

Claims

1. A method for constructing a sea surface wind field inversion model based on a dual-frequency wind field radar, characterized in that: The sea surface wind field inversion model includes the sea surface wind speed inversion model and the geophysical model function, which specifically includes the following steps: Step 1: Acquire scatterometry data, microwave radiometer data and sea surface wind field data; Based on the latitude, longitude and time information of the scatterometry data, use the spatiotemporal matching threshold method to quasi-synchronously match the microwave radiometer data and sea surface wind field data with the scatterometry data to obtain active and passive data sets; The scattering measurement data include gridded C-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth and Ku-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth; Each grid includes a plurality of C-band sea surface backscatter coefficient data and their corresponding scattering measurement azimuths and a plurality of Ku-band sea surface backscatter coefficient data and their corresponding scattering measurement azimuths; The microwave radiometer data includes C-band sea surface brightness temperature data, X-band sea surface brightness temperature data and Ka-band sea surface brightness temperature data; The sea surface wind field data includes sea surface wind speed and sea surface wind direction; Step 2: Based on the active and passive data sets, the relationship between the scattering measurement data and the sea surface wind field data is constructed, that is, the geophysical model function σ is obtained. 0，gmf ; in: Ws represents the sea surface wind speed, m / s; Relative wind direction, which is the angle between the sea surface wind direction and the scattering measurement azimuth, °; M represents the active and passive spectrum gradient ratio; σ 0,Cmeas Represents the sea surface backscatter coefficient data in the C band; σ 0,Kumeas Represents the Ku-band sea surface backscatter coefficient data; TB C Represents the sea surface brightness temperature data of C band, K; TB Ka Represents the sea surface brightness temperature data in the Ka band, K; a0, a1 and a2 are the coefficients of the geophysical model function σ0; Step 3: Use the maximum likelihood estimation method to search in the two-dimensional space of wind speed and wind direction for the wind speed and wind direction that make the cost function J reach the local minimum as the fuzzy solution of the current grid, obtain multiple fuzzy solutions corresponding to all grids, and use the circle median filtering method to find the optimal solution of wind speed and wind direction among multiple fuzzy solutions, that is, obtain the initial value of sea surface wind direction and sea surface wind speed inversion; in: σ 0i,Cmeas Represents the sea surface backscatter coefficient data of the i-th C band in the current grid; σ 0j,Kumeas Represents the sea surface backscatter coefficient data of the jth Ku band in the current grid; σ 0i,Cgmf represents the geophysical mode value of the geophysical model function of the i-th C band in the current grid; σ 0j,Kugmf represents the geophysical mode value of the geophysical model function of the jth Ku band in the current grid; Var(σ 0j,Kumeas ) represents the variance of the sea surface backscatter coefficient data of the i-th C band in the current grid; Var(σ 0j,Kumeas ) represents the variance of the j-th Ku-band sea surface backscatter coefficient data in the current grid; i represents the serial number of the C band that falls into the current grid; j represents the serial number of the Ku band falling into the current grid; Step 4: Using the microwave radiometer data in the active and passive data sets obtained in step 1 and the initial value of the sea surface wind speed inversion obtained in step 3 as input, and using the sea surface wind speed in the active and passive data sets obtained in step 1 as output, the deep learning network model is trained to obtain a sea surface wind speed inversion model.

2. The method for constructing a sea surface wind field inversion model based on a dual-frequency wind field radar as claimed in claim 1, characterized in that: The sea surface wind speed inversion model includes an input layer, a hidden layer and an output layer; The input layer is used to input parameters, which include microwave radiometer data and initial values ​​of sea surface wind speed inversion; The hidden layer is used to obtain the corresponding relationship between the parameter and the sea surface wind speed data through activation function mapping; The output layer is used to output the sea surface wind speed.

3. A method for active and passive joint inversion of high wind speed sea surface wind field based on dual-frequency wind field radar, characterized in that: The specific steps include: S1, collects active and passive data sets of the sea surface; Collect scattering measurement data, microwave radiometer data and sea surface wind field data; based on the longitude, latitude and time information of the scattering measurement data, use the spatiotemporal matching threshold method to quasi-synchronously match the microwave radiometer data and sea surface wind field data with the scattering measurement data to obtain the active and passive data sets of the sea surface; The scattering measurement data include gridded C-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth and Ku-band sea surface backscattering coefficient data and its corresponding scattering measurement azimuth; Each grid includes a plurality of C-band sea surface backscatter coefficient data and their corresponding scattering measurement azimuths and a plurality of Ku-band sea surface backscatter coefficient data and their corresponding scattering measurement azimuths; The microwave radiometer data includes C-band sea surface brightness temperature data, X-band sea surface brightness temperature data and Ka-band sea surface brightness temperature data; The sea surface wind field data includes sea surface wind speed and sea surface wind direction; S2, traverse the values ​​in the two-dimensional space of wind speed and wind direction, input the values ​​into the geophysical model function obtained in step 2 of the method for constructing a sea surface wind field inversion model based on a dual-frequency wind field radar according to claim 1 or 2, obtain the geophysical model value, input it and the sea surface active and passive data set obtained in S1 into the cost function J, and obtain the initial inversion values ​​of sea surface wind direction and sea surface wind speed; S3, input the microwave radiometer data in the sea surface active and passive data set obtained by S1 and the initial value of sea surface wind speed inversion obtained by S2 into the sea surface wind speed inversion model constructed by the method for constructing a sea surface wind field inversion model based on dual-frequency wind field radar as described in claim 1 or 2 for inversion to obtain the sea surface wind speed.

Images