Ocean wind wave and current integrated inversion method based on loop scanning SAR (Synthetic Aperture Radar)

Through the integrated inversion method of marine wind and wave flow based on ring-sweep SAR, the combined inversion of wind field, flow field and wave spectrum is used to perform joint inversion of wind field, flow field and wave spectrum, and the information in the prior art is solved The problems of hysteresis and assimilation errors realize high-precision wind and wave flow inversion and real-time data acquisition.

CN118962680APending Publication Date: 2024-11-15DALIAN MARITIME UNIVERSITY
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411257712.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing marine remote sensing technology is difficult to achieve high-precision synchronous acquisition and inversion of wind, wave and flow information, resulting in information lag and assimilation errors, and it is impossible to effectively utilize the coupling relationship between wind, wave and flow to improve the inversion accuracy.

Method used

Using the integrated inversion method of ocean wind and wave flow based on ring-sweep SAR, by obtaining initial guessing wind field information and SAR images, inversion is carried out using scattering intensity, Doppler radial velocity and image spectral model to construct the overall cost function to realize Joint inversion of wind field, flow field and wave spectrum.

Benefits of technology

The simultaneous inversion of wind and wave flow on the ocean surface is achieved, which improves the inversion accuracy, especially in the estimation accuracy of wind direction and average wave period, which reduces errors and meets the real-time data needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118962680A_ABST
    Figure CN118962680A_ABST
Patent Text Reader

Abstract

The invention discloses an ocean wind wave and flow integrated inversion method based on a circular scanning SAR, and the method comprises the following steps: carrying out the wind field inversion based on a circular scanning SAR scattering intensity forward modeling model, and obtaining the corrected wind field information and a corrected scattering intensity cost function; the corrected wind field information is input into a flow field inversion model, flow field inversion is carried out, and a Doppler migration cost function is obtained through a Doppler radial velocity forward modeling model; inputting the corrected wind field information into a wave spectrum inversion model, carrying out wave spectrum inversion, and obtaining an image spectrum cost function based on an SAR image through a loop scanning SAR image spectrum forward modeling model; constructing an overall cost function based on the corrected scattering intensity cost function, the Purler offset cost function and the image spectrum cost function; joint inversion of the wind field, the flow field and the wave spectrum is carried out on the corrected wind field information again based on the overall cost function, and the minimum value of the overall cost function is obtained, namely, the optimal flow field information, the optimal wave spectrum information and the optimal wind field information are obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of fully automated products and relates to an integrated inversion method for ocean wind, wave and current based on circular scanning SAR. Background Art

[0002] With the development of remote sensing methods, existing sensors have a certain ability to obtain information about wind, waves, and currents on the ocean surface, but there are still certain defects in the information acquisition ability. It is difficult to achieve high-precision synchronous acquisition and inversion of wind, wave, and current information on the ocean surface. Usually, it is necessary to rely on assimilation and reanalysis technology to assimilate data from different sources, different resolutions, and different time and space coverage to obtain wind, wave, and current information of the same time and space grid. On the one hand, this leads to information lag, and it is often difficult to meet the requirements of situations that require real-time data such as ocean meteorological information guarantee; on the other hand, due to the asynchronous observation time and space of each data source and the different resolutions during the assimilation process, for fast-changing ocean elements such as wind, waves, and currents, assimilation errors are inevitable.

[0003] In addition, in radar ocean remote sensing measurements, the three elements of ocean surface wind, waves, and currents are closely coupled. For example, the wind and wave part of the wave spectrum is directly determined by the local wind. In flow field measurements, the measured Doppler velocity also includes interference items such as Bragg wave phase velocity, wave orbit velocity, wind-induced current, and Stokes drift. Among them, Bragg wave phase velocity, wave orbit velocity, etc. are all related to the wave spectrum, and wind-induced current is directly related to wind speed. Wind field inversion relies on measuring the sea surface scattering intensity for inversion, and the sea surface scattering intensity actually depends on the relative flow velocity between the wind field and the sea surface flow field, rather than the absolute wind speed. When only single-element information can be obtained, this coupling relationship is the error term that restricts the inversion accuracy. If the three elements of information can be obtained synchronously, this coupling relationship will be transformed into a mutually corroborated relationship, which will not only not restrict the inversion accuracy, but will help to further improve the inversion accuracy as a whole.

[0004] To fill the gap in the observation of wind, flow and wave comprehensive measurements, scholars have proposed several innovative ocean microwave remote sensing satellite schemes, such as the Sea Surface Kinematics Multiscale Monitoring Satellite Mission (SKIM). [1] , Winds and Currents Mission (WaCM) [2] and Seastar [3]missions. These missions propose to provide various simultaneous measurements of ocean surface vector flow fields, wave spectra, and ocean surface vector wind fields. Although these missions all measure surface vector flow fields, Stokes drift, wave spectra, and sea surface vector wind fields simultaneously, each mission is different in nature. WaCM focuses on global monitoring of flow and wind fields using scatterometers to seek better observations of wind-flow interactions and their impacts on global surface fluxes. SKIM's goals include measuring total vector flow fields and wave directional spectra to explore strong currents and associated extreme waves. Seastar focuses on high-precision observations of total ocean surface vector flow fields and vector wind fields. Unfortunately, none of these three missions have the capability to simultaneously observe wind, wave, and flow fields.

[0005] Circular scanning synthetic aperture radar (Synthetic aperture radar, SAR) [4] It not only has the multi-angle observation capability of the ring-scanning scatterometer and spectrometer, but also has the unique two-dimensional high-resolution imaging capability of SAR, which can simultaneously obtain wind field, wave spectrum, and flow field information, and can make full use of the coupling characteristics between them to perform integrated inversion, thereby improving the overall inversion accuracy. This makes the ring-scanning SAR a strong candidate for the new generation of ocean remote sensing sensors. Summary of the invention

[0006] In order to solve the above problems, the technical solution adopted by the present invention is: an integrated inversion method of ocean wind, wave and current based on circular scanning SAR, comprising the following steps:

[0007] Obtain preliminary wind field information and SAR images;

[0008] Taking the initial guess of wind field as the initial input information, the wind field inversion is performed based on the forward model of the scattering intensity of the circular scanning SAR to obtain the corrected wind field information and the corrected scattering intensity cost function;

[0009] The corrected wind field information is input into the flow field inversion model to perform flow field inversion, and the Doppler shift cost function is obtained through the Doppler radial velocity forward model.

[0010] The corrected wind field information is input into the wave spectrum inversion model to perform wave spectrum inversion. The image spectrum cost function is obtained based on the SAR image through the forward model of the ring-scanning SAR image spectrum.

[0011] An overall cost function is constructed based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function;

[0012] Based on the overall cost function, the wind field, flow field and wave spectrum are jointly inverted again. When the cost function is set to be less than the threshold, the iterative cycle is terminated to obtain the minimum value of the overall cost function, that is, to obtain the optimal flow field information, wave spectrum information and wind field information.

[0013] Furthermore: the forward model of the scattering intensity of the circular scanning SAR is mainly based on the Ku band, and uses an empirical model based on KuMod;

[0014] The scattering intensity is usually expressed as normalized radar cross section. The GMF of Ku-band NRCS is KuMod, and NRCS is expressed as:

[0015] σ 0 ~KuMOD(φ,|u 10 -c,θ,pol)

[0016] Among them, φ refers to the angle of the relative wind direction to the radar azimuth, that is, u 10 -c is the angle between the direction of the antenna and the azimuth angle. φ is 0°, indicating the direction against the wind, that is, the wind blows towards the radar antenna. θ is the incident angle. pol is the polarization mode, VV or HH. u 10 represents the vector wind field at the upper 10 meters of the ocean surface, c represents the vector flow field, and pol is the polarization mode;

[0017] Therefore, the corrected scattering intensity cost function is expressed as:

[0018]

[0019] in, is the NRCS observed at the i-th azimuth, KuMOD(φ,|u 10 -c|,θ,pol) is the NRCS calculated by the NRCS geophysical model, δ σ is the mean square error of NRCS.

[0020] Further: The Doppler radial velocity forward model KuDOP is obtained as follows:

[0021] The Doppler shift caused by the wind field is f KuDOP It is expressed as:

[0022] f KuDOP =KuDOP(φ,|u 10 -c,θ,pol)

[0023] The Doppler shift caused by the flow field is f c It is expressed as:

[0024]

[0025] in, is the radar incident direction unit vector, λ is the radar wavelength;

[0026] The total Doppler frequency deviation observed by circular scanning SAR is the sum of the two, namely:

[0027] f dc =f KuDOP +f c

[0028] From the estimated Doppler shift f dc Subtract f from KuDOP , to accurately estimate f c .

[0029] Therefore, the Doppler shift cost function is expressed as:

[0030]

[0031] Among them, f dc,i is the Doppler frequency deviation observed at the i-th azimuth, KuDOP(φ,|u 10 -c|,θ,pol) is the Doppler frequency shift caused by the flow field calculated based on the geophysical model, δ c is the mean square error of the Doppler shift.

[0032] Further: the process of obtaining the forward model of the circular scanning SAR image spectrum is as follows:

[0033] According to the PARSA algorithm, for each wave system, a random model is established to define the probability of different wave propagation directions, wavelengths or energies deviating from prior knowledge. The specified probability quantifies the confidence of the initial guess of the wave spectrum, and the overall shape of the spectrum is obtained using a numerical wave model.

[0034] The SAR information is then used to adjust the energy, propagation direction, wavelength, or directional spread parameters to produce deviations from the a priori spectrum;

[0035] Assume that there is only one partition of the spectrum, B, which can be expressed in polar grid (φ, ρ) as:

[0036]

[0037] Among them, X φ and X Δφ They are used to determine the direction of wave propagation and the width of the directional spectrum, X k Adjust the phase velocity of the main wave, X E It is equivalent to the effective wave height, φ0 is the peak direction, ρ is the polar diameter of the polar coordinates, and φ is the polar angle of the polar coordinates. At this time, the initial guess of the wave spectrum s is expressed as:

[0038] s=B

[0039] Simulation Image Spectrum Calculate the image spectrum P that represents the "real" K , considering different azimuth angles, P i It is expressed as:

[0040]

[0041] where α1 is a value with unit mean and standard deviation Gaussian distribution variable, describing the error of the overall energy level of the wave spectrum, α2 is zero mean and standard deviation Gaussian distribution variable, describing the uncertainty of the cutoff wavelength of the forward model of the circular scanning SAR image spectrum. x is the distance wave number, R is the radius from the satellite to the earth, V is the satellite flight speed, represents the i-th azimuth;

[0042] Therefore, the image spectrum cost function of wave spectrum inversion can be expressed as:

[0043]

[0044] in, is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, P i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, δ s is the mean square error of the image spectrum.

[0045] Further: the expression of the overall cost function is as follows:

[0046]

[0047] in, f dc,i are the NRCS and Doppler velocity observed at the i-th azimuth, P i,j is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, KuMOD(φ,|u 10 -c|,θ,pol), KuDOP(φ,|u 10 -c|,θ,pol) are the geophysical models of Doppler frequency shift caused by NRCS and flow field, respectively, i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, u 10 , c and s are wind field vector, flow field vector and wave spectrum respectively, δ σ , δ c , δ sare NRCS, Doppler shift and mean square error of image spectrum respectively.

[0048] An integrated ocean wind, wave and current inversion device based on circular scanning SAR, comprising:

[0049] Acquisition module: used to obtain preliminary wind field information and SAR images;

[0050] Wind field inversion module: used to take the initial guess of the wind field as the initial input information, perform wind field inversion based on the forward model of the scattering intensity of the circular scanning SAR, and obtain the corrected wind field information and the corrected scattering intensity cost function;

[0051] Flow field inversion module: used to input the corrected wind field information into the flow field inversion model, perform flow field inversion, and obtain the Doppler shift cost function through the Doppler radial velocity forward model;

[0052] Wave spectrum inversion module: used to input the corrected wind field information into the wave spectrum inversion model, perform wave spectrum inversion, and obtain the image spectrum cost function based on the SAR image through the forward model of the ring-scanning SAR image spectrum;

[0053] Overall function construction module: used to construct an overall cost function based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function;

[0054] Joint inversion module: used to perform joint inversion of wind field, flow field and wave spectrum again based on the overall cost function. When the cost function is set to be less than the threshold, the iterative loop ends and the minimum value of the overall cost function is obtained, that is, the optimal flow field information, wave spectrum information and wind field information are obtained.

[0055] A readable storage medium stores a program module, wherein the program module is executed in a processor to implement any of the methods described above.

[0056] The invention provides an integrated inversion method for ocean wind, wave and current based on circular scanning SAR, which has the following advantages: the method comprehensively considers the errors between three inverted physical quantities and measured values. The measured values ​​include Normalized Radar Cross Section (NRCS or σ 0 ), Doppler frequency shift and SAR image spectrum. These three measurements have direct or indirect relationships with the three physical quantities of wind wave current. The wind wave current obtained by inversion can be used to calculate the NRCS KuMOD (φ,u n -c,θ,pol), Doppler frequency shift KuDOP(φ,u 10 -c,θ,pol) and image spectrum P i,j (s(u 10 -c)), NRCSσ can also be extracted from the radar signali , Doppler frequency shift f dc,i and image spectrum The optimal inversion value is considered to be the one with the smallest error between the result calculated by the overall cost function and the result extracted by the radar signal.

[0057] This method can not only realize the simultaneous inversion of ocean surface wind, waves and currents, but also significantly improve the estimation accuracy of several other hydrological parameters at any azimuth interval angle, except wind direction and mean wave period, compared with the separate inversion. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0059] Figure 1 It is a schematic diagram of circular scanning SAR;

[0060] Figure 2 This is a flow chart of the integrated inversion method of ocean wind, wave and current based on circular scanning SAR; DETAILED DESCRIPTION

[0061] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0062] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Circular scanning synthetic aperture radar (SAR) is a sensor with great potential in ocean remote sensing. It can realize the synchronous monitoring of wide-area ocean wind, waves and currents. It is a strong candidate for the new generation of ocean remote sensing sensors. The schematic diagram of circular scanning SAR is shown in Figure 1 shown.

[0064] Figure 2This is a flow chart of the integrated inversion method of ocean wind, wave and current based on circular scanning SAR;

[0065] An integrated inversion method for ocean wind, wave and current based on circular scanning SAR comprises the following steps:

[0066] S1: Obtain preliminary wind field information and SAR images;

[0067] S2: Taking the initial guess of the wind field as the initial input information, based on the forward model of the scattering intensity of the ring-scanning SAR, the wind field is inverted to obtain the corrected wind field information and the corrected scattering intensity cost function;

[0068] S3: inputting the corrected wind field information into the flow field inversion model, performing flow field inversion, and obtaining the Doppler shift cost function through the Doppler radial velocity forward model;

[0069] S4: inputting the corrected wind field information into the wave spectrum inversion model to perform wave spectrum inversion, and obtaining the image spectrum cost function based on the SAR image through the ring scanning SAR image spectrum forward model;

[0070] S5: constructing an overall cost function based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function;

[0071] S6: Based on the overall cost function, the wind field, flow field and wave spectrum are jointly inverted again for the corrected wind field information. When the cost function is set to be less than the threshold, the iterative loop is terminated, which means that the minimum value of the overall cost function is obtained at this time.

[0072] The threshold is 0.001.

[0073] Steps S1 / S2 are executed sequentially, and then S3 and S4 are executed in parallel. After S3 / S4 is executed, S5 and S6 are executed sequentially;

[0074] The forward model of circular scanning SAR observations includes the circular scanning SAR scattering intensity forward model, the Doppler radial velocity forward model and the circular scanning SAR image spectrum forward model.

[0075] In the following expression, the vector wind field at the upper 10 m above the ocean surface is represented by u 10 , wind speed (wspd)u 10 is understood as the wind moving relative to the ocean surface, c is the ocean surface vector flow field, and the relative wind vector relative to the sea surface is represented by u 10 -c.

[0076] First, the forward model of wind wave current scattering intensity NRCS, Doppler frequency deviation and image spectrum observation is established. In the following expression, the vector wind field at the upper 10 meters of the ocean surface is expressed as u 10, wind speed (wspd)u 10 is understood as the wind moving relative to the ocean surface, c is the ocean surface vector flow field, and the relative wind vector relative to the sea surface is represented by u 10 -c.

[0077] The construction process of the circular scanning SAR scattering intensity forward model is as follows:

[0078] There are two main types of forward models for scattering intensity: one is theoretical simulation methods, such as dual-scale models, three-scale models, combined surface models, etc. These methods have a wide range of applications but large errors; the other is empirical models, typically including a series of geophysical model functions (GMF) in different bands such as LMOD, CMOD, XMOD, and KuMOD. These models have high accuracy but a narrow scope of application, and are only applicable to specific bands, viewing angles, polarizations, etc. When studying the mapping relationship between microwave scattering intensity and wind field, they are usually based on GMF.

[0079] This application mainly focuses on the Ku band, so the empirical model based on KuMod is used. The scattering intensity is usually expressed in terms of the Normalized Radar Cross Section (NRCS or σ 0 )express,

[0080] In this application, the GMF of Ku-band NRCS is taken from NSCAT and is called KuMod. 0 There is a dependency on the incident angle, wind speed and wind direction φ. When the radar line of sight is consistent with the wind direction (headwind and tailwind), σ 0 Strongest, when the antenna line of sight is perpendicular to the wind direction (crosswind), σ 0 The weakest. For incident angles above 25°, σ under different polarizations 0 The same changes are shown with wind speed, indicating that at smaller incident angles (i.e., closer to the subsatellite point), the impact on various polarization modes is not significant. 0 It decreases rapidly with increasing incident angle, especially for HH polarization. Therefore, NRCS can be expressed as:

[0081] σ 0 ~KuMOD(φ,u 10 -c,θ,pol)(1)

[0082] Among them, φ refers to the angle of the relative wind direction to the radar azimuth, that is, u 10 -c is the angle between the direction of the antenna and the azimuth angle. φ is 0°, indicating the direction against the wind, that is, the wind blows towards the radar antenna. θ is the incident angle. pol is the polarization mode, VV or HH. u10 represents the vector wind field at the upper 10 meters above the ocean surface, c represents the vector flow field, and pol represents the polarization mode.

[0083] Therefore, the scattering intensity cost function of wind field inversion can be expressed as:

[0084]

[0085] in, is the NRCS observed at the i-th azimuth, KuMOD(φ,|u 10 -c|,θ,pol) is the NRCS calculated by the geophysical model of NRCS, δ σ is the mean square error of NRCS.

[0086] However, it should be noted that this GMF was developed for scatterometers and represents σ 0 Averaged over the typical spatial scale of a scatterometer, i.e., tens of kilometers. At moderate angles of incidence, other geophysical effects can also affect σ 0 , such as flow field, water depth measurement or breaking waves will affect σ 0 This has a minor impact and will not be discussed here.

[0087] The construction process of the Doppler radial velocity forward model is as follows:

[0088] The empirical model for estimating Doppler frequency shift using the Doppler radial velocity forward model is based on the Envisat C-band CDOP model. The CDOP model is derived from the global data set of ECMWF vector wind field observations with Envisat / ASAR, and provides an expression for the average Doppler frequency shift in the C-band with wind speed, wind direction and incident angle.

[0089] Therefore, as a global average, CDOP represents the effect of wind and waves on the Doppler shift. With this model, CDOP includes the effect of waves on the Doppler shift from the full wave spectrum, but secondary effects such as swell and wind-wave-current interactions are not included in this statistically averaged relationship.

[0090] The KuDop empirical model is obtained by frequency scaling (about 13.6 / 5.3=2.6) of the CDOP model to adapt it to the Ku band.

[0091] The Doppler shift caused by wind and waves is expressed as the equivalent ocean surface flow velocity after projecting from the line of sight (LOS) slant range to the sea surface, which is called the wind-wave induced artifact surface velocity (WASV). WASV is dependent on the angle of incidence, wind speed, and azimuth. WASV is greatly affected by wind speed, and its magnitude can reach 0.5m / s to 2m / s, which is as large or larger than the contribution of most ocean surface flow fields to Doppler shift. The impact of the angle of incidence on WASV is relatively small. WASV decreases with increasing angle of incidence. WASV is most strongly affected by relative wind direction. In the downwind direction, there is a clear difference between the two polarizations, and unlike NRCS, this difference will appear even at angles of incidence within 25°. However, since the surface flow field needs to be projected onto the radar slant range, the total Doppler sensitivity to the ocean surface flow field will be higher at larger angles of incidence. For these reasons, Doppler frequency observations at higher incidence angles appear to be a better choice.

[0092] Then, the Doppler shift caused by the wind field is f KuDOP It can be expressed as:

[0093] f KuDOP =KuDOP(φ,u 10 -c,θ,pol) (3)

[0094] The Doppler shift caused by the flow field itself can be expressed as:

[0095]

[0096] in, is the radar incident direction unit vector, λ is the radar wavelength; the radar incident direction vectors in different viewing directions are different.

[0097] The total Doppler frequency deviation observed by circular scanning SAR is the sum of the two, namely:

[0098] f dc =f KuDOP +f c (5)

[0099] Therefore, it is necessary to estimate the Doppler frequency shift f dc Subtract f from KuDOP , we can accurately estimate f c .

[0100] The process of obtaining the forward model of the circular scanning SAR image spectrum is as follows:

[0101] Therefore, the Doppler shift cost function of flow field inversion can be expressed as:

[0102]

[0103] Among them, f dc,i is the Doppler frequency deviation observed at the i-th azimuth, KuDOP(φ,|u 10 -c|,θ,pol) is the Doppler frequency shift caused by the flow field calculated based on the geophysical model, δ c is the mean square error of the Doppler shift.

[0104] The wave spectrum can be used to describe the distribution of wave energy at wave frequencies, and this information can be used to estimate the wave shape and wave-induced motion. Therefore, it can be used for numerical simulation and calculation of the ocean surface. In 1991, Hasselmann et al. gave a closed nonlinear integral transformation from the wave spectrum to the SAR image spectrum, namely the SAR image forward model, and proposed an inversion method that requires prior information on the waves. Because it was developed at the Max Planck Institute in Germany, it is also called the MPI method. The nonlinear mapping from the wave spectrum s to the SAR image spectrum P can be expressed as:

[0105]

[0106] Among them, i is the corresponding i-th azimuth.

[0107] According to the PARSA (Partition Rescale And Shift Algorithm) algorithm, for each wave system, a random model is established to define the probability that the propagation direction, wavelength or energy of different waves deviate from the prior knowledge. The specified probability quantifies the confidence of the initial guess of the wave spectrum. The overall shape of the spectrum is obtained using a numerical wave model, and then the parameters such as energy, propagation direction, wavelength or directional extension are adjusted using SAR information to produce deviations from the prior spectrum. Assuming that there is only one partitioned spectrum, namely B, it can be expressed in the polar coordinate grid (φ, ρ) as:

[0108]

[0109] Among them, X φ and X Δφ They are used to determine the direction of wave propagation and the width of the directional spectrum respectively. k Adjust the phase velocity of the main wave, X E It is equivalent to the effective wave height, which determines the total energy of the wave spectrum. φ0 is the peak direction, ρ is the polar diameter of the polar coordinates, and φ is the polar angle of the polar coordinates. At this time, the initial guess of the wave spectrum can be expressed as:

[0110] s=B (9)

[0111] Since the image spectrum can also be considered as the cross spectrum with sub-aperture interval time of 0, the simulated image spectrum The “real” image spectrum P can be calculated K Considering different azimuth angles, P i It can be expressed as:

[0112]

[0113] where α1 is a value with unit mean and standard deviation A Gaussian distribution variable that describes the error in the overall energy level of the wave spectrum. α2 is zero mean and standard deviation It is a Gaussian distribution variable that describes the uncertainty of the cutoff wavelength of the forward model of the circular scanning SAR image spectrum. x is the distance wave number, R is the radius from the satellite to the earth, V is the satellite flight speed, represents the i-th azimuth.

[0114] Therefore, the image spectrum cost function of wave spectrum inversion can be expressed as:

[0115]

[0116] in, is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, P i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, δ s is the mean square error of the image spectrum.

[0117] According to the above three forward models, the cost function involved must include the above three physical quantities, so the overall cost function is:

[0118]

[0119] in, f c,i are the NRCS and Doppler velocity observed at the i-th azimuth, P i,j is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, KuMOD(φ,|u 10 -c|,θ,pol), KuDOP(φ,|u 10 -c|,θ,pol) are the geophysical models of Doppler frequency shift caused by NRCS and flow field, respectively, i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, u 10 , c and s are wind field vector, flow field vector and wave spectrum respectively, δσ , δ c , δ s They are NRCS, Doppler shift and mean square error of image spectrum. These three factors normalize the cost function, so that the three errors of wind, waves and current have the same importance in the overall cost function.

[0120] The vector wind field u obtained by inversion 10 , flow field c and wave spectrum s are:

[0121] [u 10 ,c,s]=argmin(J(u 10 ,c,s)) (13)

[0122] Since the wind field not only affects the accuracy of flow field inversion, but is also important prior information for wave spectrum inversion, this method sets a preliminary guess wind field as the initial input information. After optimizing and iterating the preliminary guess wind field, not only can the cost function related to NRCS be obtained, but also the cost functions of Doppler velocity and image spectrum can be obtained after being brought into flow field inversion and wave spectrum inversion. After adding these three cost functions, the total cost function is iteratively corrected. When the cost function is set to be less than 0.001, the iterative cycle is terminated, that is, it is considered that the minimum value of the overall cost function is obtained at this time, and the optimal flow field information, wave spectrum information and wind field information can be obtained.

[0123] An integrated ocean wind, wave and current inversion device based on circular scanning SAR, comprising:

[0124] Acquisition module: used to obtain preliminary wind field information and SAR images;

[0125] Wind field inversion module: used to take the initial guess of the wind field as the initial input information, perform wind field inversion based on the forward model of the scattering intensity of the circular scanning SAR, and obtain the corrected wind field information and the corrected scattering intensity cost function;

[0126] Flow field inversion module: used to input the corrected wind field information into the flow field inversion model, perform flow field inversion, and obtain the Doppler shift cost function through the Doppler radial velocity forward model;

[0127] Wave spectrum inversion module: used to input the corrected wind field information into the wave spectrum inversion model, perform wave spectrum inversion, and obtain the image spectrum cost function based on the SAR image through the forward model of the ring-scanning SAR image spectrum;

[0128] Overall function construction module: used to construct an overall cost function based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function;

[0129] Joint inversion module: used to perform joint inversion of wind field, flow field and wave spectrum again based on the overall cost function. When the cost function is set to be less than the threshold, the iterative loop ends and the minimum value of the overall cost function is obtained, that is, the optimal flow field information, wave spectrum information and wind field information are obtained.

[0130] 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. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0131] [1]Ardhuin F, Aksenov Y, Benetazzo A, et al. Measuring currents, icedrift, and waves from space: the Sea surface KInematics Multiscale monitoring (SKIM) concept [J]. Ocean Science, 2018, 14(3): 337–354.

[0132] [2] Rodriguez E, Bourassa MA, Chelton D, et al. The Winds and CurrentsMission Concept[J]. Frontiers in Marine Science, 2019.

[0133] [3]Gommenginger C,Chapron B,Hogg A,et al.SEASTAR:A Mission to StudyOcean Submesoscale Dynamics and Small-Scale Atmosphere-Ocean Processes inCoastal,Shelfand Polar Seas[J].Frontiers inMarine Science,2019,6:457.

[0134] [4] Sun Bing, Zhou Yinqing, Li Tianchi, et al. Fast focusing imaging algorithm for circular scanning SAR[J]. Journal of Beijing University of Aeronautics and Astronautics, 2007, 33(7):4.

Claims

1. An integrated inversion method for ocean wind, wave and current based on circular scanning SAR, Features: The following steps are involved: Obtain preliminary wind field information and SAR images; Taking the initial guess of wind field as the initial input information, the wind field inversion is performed based on the forward model of the scattering intensity of the circular scanning SAR to obtain the corrected wind field information and the corrected scattering intensity cost function; The corrected wind field information is input into the flow field inversion model to perform flow field inversion, and the Doppler shift cost function is obtained through the Doppler radial velocity forward model. The corrected wind field information is input into the wave spectrum inversion model to perform wave spectrum inversion. The image spectrum cost function is obtained based on the SAR image through the forward model of the ring-scanning SAR image spectrum. An overall cost function is constructed based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function; Based on the overall cost function, the wind field, flow field and wave spectrum are jointly inverted again. When the cost function is set to be less than the threshold, the iterative cycle is terminated to obtain the minimum value of the overall cost function, that is, to obtain the optimal flow field information, wave spectrum information and wind field information.

2. The method for integrated ocean wind, wave and current inversion based on circular scanning SAR according to claim 1, Features: The forward model of the scattering intensity of the circular scanning SAR is mainly based on the Ku band, and uses an empirical model based on KuMod; The scattering intensity is usually expressed as normalized radar cross section. The GMF of Ku-band NRCS is KuMod, and NRCS is expressed as: σ 0 ~ForMOD(φ,|u 10 −c|,θ,pol) Among them, φ refers to the angle of the relative wind direction to the radar azimuth, that is, u 10 -c is the angle between the direction of the antenna and the azimuth angle. φ is 0°, indicating the direction against the wind, that is, the wind blows towards the radar antenna. θ is the incident angle. pol is the polarization mode, VV or HH. u 10 represents the vector wind field at the upper 10 meters of the ocean surface, c represents the vector flow field, and pol is the polarization mode; Therefore, the corrected scattering intensity cost function is expressed as: in, is the NRCS observed at the i-th azimuth, KuMOD(φ,|u n -c|,θ,pol) is the NRCS calculated by the NRCS geophysical model, δ σ is the mean square error of NRCS.

3. The integrated inversion method of ocean wind, wave and current based on circular scanning SAR according to claim 1, Features: The Doppler radial velocity forward model KuDOP is obtained as follows: The Doppler shift caused by the wind field is f KuDOP It is expressed as: f KuDOP =ForDOP(φ,|u 10 −c|,θ,pol) The Doppler shift caused by the flow field is f c It is expressed as: in, is the radar incident direction unit vector, λ is the radar wavelength; The total Doppler frequency deviation observed by circular scanning SAR is the sum of the two, namely: f dc =f KuDOP +f c From the estimated Doppler shift f dc Subtract f from KuDOP , to accurately estimate f c . Therefore, the Doppler shift cost function is expressed as: Among them, f dc,i is the Doppler frequency deviation observed at the i-th azimuth, KuDOP(φ,|u 10 -c|,θ,pol) is the Doppler frequency shift caused by the flow field calculated based on the geophysical model, δ c is the mean square error of the Doppler shift.

4. The method for integrated ocean wind, wave and current inversion based on circular scanning SAR according to claim 1, Features: The process of obtaining the forward model of the circular scanning SAR image spectrum is as follows: According to the PARSA algorithm, for each wave system, a random model is established to define the probability of different wave propagation directions, wavelengths or energies deviating from prior knowledge. The specified probability quantifies the confidence of the initial guess of the wave spectrum, and the overall shape of the spectrum is obtained using a numerical wave model. The SAR information is then used to adjust the energy, propagation direction, wavelength, or directional spread parameters to produce deviations from the a priori spectrum; Assume that there is only one partition of the spectrum, B, which can be expressed in polar grid (φ, ρ) as: Among them, X φ and X Δφ They are used to determine the direction of wave propagation and the width of the directional spectrum, X k Adjust the phase velocity of the main wave, X E Equivalent to the effective wave height, φ 0 is the peak direction, ρ is the polar diameter of the polar coordinates, and φ is the polar angle of the polar coordinates. At this time, the initial guess of the wave spectrum s is expressed as: s=B Simulation Image Spectrum Calculate the image spectrum P that represents the "real" K , considering different azimuth angles, P i It is expressed as: Among them, α 1 is a value with unit mean and standard deviation Gaussian distribution variable, describing the error of the overall energy level of the wave spectrum, α 2 is zero mean and standard deviation Gaussian distribution variable, describing the uncertainty of the cutoff wavelength of the forward model of the circular scanning SAR image spectrum. x is the distance wave number, R is the radius from the satellite to the earth, V is the satellite flight speed, represents the i-th azimuth; Therefore, the image spectrum cost function of wave spectrum inversion is expressed as: in, is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, P i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, δ s is the mean square error of the image spectrum.

5. The integrated inversion method of ocean wind, wave and current based on circular scanning SAR according to claim 1, Features: The expression of the overall cost function is as follows: in, f dc,i are the NRCS and Doppler velocity observed at the i-th azimuth, P i,j is the image spectrum observed at the jth spectrum point obtained by inversion at the i-th azimuth angle, KuMOD(φ,|u 10 -c|,θ,pol), KuDOP(φ,|u 10 -c|,θ,pol) are the geophysical models of Doppler frequency shift caused by NRCS and flow field, respectively, i,j (s(u 10 -c)) is the image spectrum obtained by substituting the wave spectrum s obtained by inverting the initial guess spectrum into the forward model, u 10 , c and s are wind field vector, flow field vector and wave spectrum respectively, δ σ ,δ c ,δ s are NRCS, Doppler shift and mean square error of image spectrum respectively.

6. An integrated ocean wind, wave and current inversion device based on circular scanning SAR, Features: Including acquisition module: used to obtain preliminary wind field information and SAR images; Wind field inversion module: used to take the initial guess of the wind field as the initial input information, perform wind field inversion based on the forward model of the scattering intensity of the circular scanning SAR, and obtain the corrected wind field information and the corrected scattering intensity cost function; Flow field inversion module: used to input the corrected wind field information into the flow field inversion model, perform flow field inversion, and obtain the Doppler shift cost function through the Doppler radial velocity forward model; Wave spectrum inversion module: used to input the corrected wind field information into the wave spectrum inversion model, perform wave spectrum inversion, and obtain the image spectrum cost function based on the SAR image through the forward model of the ring-scanning SAR image spectrum; Overall function construction module: used to construct an overall cost function based on the corrected scattering intensity cost function, the Püller shift cost function and the image spectrum cost function; Joint inversion module: used to perform joint inversion of wind field, flow field and wave spectrum again based on the overall cost function. When the cost function is set to be less than the threshold, the iterative loop ends and the minimum value of the overall cost function is obtained, that is, the optimal flow field information, wave spectrum information and wind field information are obtained.

7. A readable storage medium storing a program module, It is characterized in that The program module is executed in a processor to implement the method according to any one of claims 1 to 5.

Citation Information

Cited By

  • Ocean internal wave flow field inversion method and device

    CN119687875A

  • Offshore wind energy short-time forecast wave effect correction method and system

    CN119740514A

  • HFR and SAR combined sea surface wind wave field parameter integrated inversion method

    CN120522709A