Satellite remote sensing in-ocean solitary wave continuous vertical structure inversion method and device

Abstract

The invention discloses an in-ocean solitary wave continuous vertical structure inversion method and device based on satellite remote sensing, and belongs to the technical field of synthetic aperture radars. The method comprises the following steps: collecting ocean real stratification data, and extracting ocean stratification parameters; numerical simulation and radar simulation experiments are carried out, an initial density field of numerical simulation is set based on ocean stratification parameters, and an inner solitary wave vertical structure-remote sensing characteristic parameter and ocean environment parameter pair sample data set is constructed; training a back propagation neural network by using the sample data set to obtain a trained internal solitary wave vertical structure inversion model; and obtaining remote sensing characteristic parameters of the target internal solitary waves and marine environment parameters of corresponding positions, namely, marine stratification parameters and sea surface wind field data, and inputting the parameters into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure. The internal solitary wave induced isodensity line offset obtained by the method has extremely high consistency with a field measurement result.

Description

Technical Field

[0001] The invention belongs to the technical field of synthetic aperture radar, and in particular relates to a method and a device for inverting the continuous vertical structure of an internal solitary wave in the ocean by satellite remote sensing. Background Art

[0002] Internal solitary waves occur in stable stratified seawater and are a type of wave that occurs inside the ocean. The presence of large-amplitude internal solitary waves can cause vertical displacements of isopycnals of tens or hundreds of meters within tens of minutes, posing a serious threat to offshore engineering, underwater communications, and submarine detection. Therefore, studying the vertical structure of internal solitary waves is crucial to ensuring the safety of offshore engineering.

[0003] The vertical structure of internal solitary waves is usually defined by the maximum displacement (i.e., amplitude) of isopycnals at different water depths. The acquisition of the vertical structure of internal solitary waves in the ocean mainly relies on field observations. However, the low observation resolution and high observation costs limit the development of field observations. Satellite observations have the advantages of all-day, high resolution, and a large range. They can clearly observe the alternating light and dark strip signals caused by internal solitary waves in the ocean. Therefore, they have application potential in the field of research on the characteristics of internal solitary waves in the ocean. However, there is no reliable method for inverting the vertical structure of internal solitary waves in a continuous layered system based on remote sensing images in the prior art. Summary of the invention

[0004] To solve the above technical problems, the present invention provides a method and device for inverting the continuous vertical structure of internal solitary waves in the ocean based on satellite remote sensing, and establishes an inversion model for the vertical structure of internal solitary waves in the ocean under a continuous layered system based on remote sensing images and not limited by the nonlinear strength of the internal solitary waves and the water depth.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] A method for inverting the continuous vertical structure of internal solitary waves in the ocean using satellite remote sensing, the method comprising:

[0007] Step 1: Collect real ocean stratification data and extract ocean stratification parameters according to the improved ocean stratification parameterization scheme;

[0008] Step 2: Conduct numerical simulation and radar simulation experiments, use the extracted ocean stratification parameters to set the initial density field of the numerical simulation, set different levels of sea surface wind speed in the radar simulation experiment, and construct a sample data set of internal solitary wave vertical structure-remote sensing characteristic parameters and ocean environment parameters;

[0009] Step 3: Use the sample data set to train the back propagation neural network to obtain a trained internal solitary wave vertical structure inversion model;

[0010] Step 4, obtaining remote sensing characteristic parameters of solitary waves within the target, ocean stratification parameters at the corresponding position, and sea surface wind field data;

[0011] Step 5: Input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind field data into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure.

[0012] In a second aspect, the present invention provides a device for inverting continuous vertical structures of internal solitary waves in the ocean by satellite remote sensing, comprising:

[0013] A stratification parameter extraction unit, used to collect real ocean stratification data and extract ocean stratification parameters according to an improved ocean stratification parameterization scheme;

[0014] The sample data set acquisition unit is used to perform numerical simulation and radar simulation experiments, use the extracted ocean stratification parameters to set the initial density field of the numerical simulation, and set different levels of sea surface wind speed in the radar simulation experiment to construct the sample data set of the vertical structure of the internal solitary wave - remote sensing characteristic parameters and ocean environment parameters;

[0015] A model acquisition unit, used for training a back propagation neural network using a sample data set to obtain a trained internal solitary wave vertical structure inversion model;

[0016] A target parameter acquisition unit is used to obtain remote sensing characteristic parameters of solitary waves in the target, ocean stratification parameters at the corresponding position, and sea surface wind field data;

[0017] The target structure acquisition unit is used to input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind field data into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure.

[0018] In a third aspect, the present invention provides an electronic device comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned satellite remote sensing method for inverting continuous vertical structures of solitary waves in the ocean.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned method for inverting the continuous vertical structure of solitary waves in the ocean using satellite remote sensing.

[0020] The beneficial effects of the present invention are:

[0021] The present invention improves the parameterization scheme of ocean stratification by adding a deep-water stratification parameter, which is more suitable for describing the real ocean stratification profile. Due to the addition of the description of deep ocean stratification, there is still a density gradient in the deep water part that supports the propagation of internal solitary waves. Therefore, the parameterized stratified profile obtained by the improved parameterization scheme can be used to describe the complete vertical structure of the internal solitary wave. The internal solitary wave numerical simulation is performed using the parameterized stratified profile as the initial density field. The internal solitary wave-induced isopycnic line offset obtained is highly consistent with the field measurement results.

[0022] The present invention constructs a high-precision internal solitary wave vertical structure inversion model in a continuous stratified system based on remote sensing images. Different from the existing theoretical models, the inversion model proposed in the present invention has a certain universality and is not limited by the strength of the nonlinearity of the internal solitary wave and the size of the water depth. In addition, the inversion model proposed in the present invention realizes the inversion of the internal solitary wave vertical structure in a continuous stratified system based on synthetic aperture radar images for the first time, gets rid of the constraints of the two-layer ocean model, and is closer to the internal solitary wave structure in the real ocean environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 The present invention is a flow chart of the method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing;

[0024] Figure 2 It is a schematic diagram comparing the floating frequency profile measured on site with the parameterized floating frequency profile of the improved parameterization scheme in the present invention;

[0025] Figure 3 It is a schematic diagram of the characteristic points of the vertical structure;

[0026] Figure 4 This is a schematic diagram of the network structure of the internal solitary wave vertical structure inversion model;

[0027] Figure 5 A Sentinel-1A SAR image and a schematic diagram of the internal solitary wave backscattering cross section;

[0028] Figure 6 A comparison diagram of the internal solitary wave-induced isopycnic line migration obtained by the parameterized floating frequency profile simulation of the present invention and the field measurement results;

[0029] Figure 7 It is a comparison diagram between the internal solitary wave vertical structure inversion model obtained based on the present invention and the field data results. DETAILED DESCRIPTION

[0030] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0031] like Figure 1As shown, it is a flow chart of the method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing of the present invention, which specifically includes the following steps:

[0032] Step 1: Collect real ocean stratification data and extract ocean stratification parameters according to the improved ocean stratification parameterization scheme. In order to invert the vertical structure of solitary waves in the real ocean environment, it is necessary to first obtain the stratification data of the real ocean environment. Collect the real ocean temperature profile data recorded by the self-sinking and floating profile detection buoy (Argo) in a certain sea area. and salinity profile data , according to formula (1) and formula (2), calculate the corresponding density profile and floating frequency profile :

[0033] (1)

[0034] (2)

[0035] in, 1028 kg / m 3 , g represents the acceleration due to gravity, It is the water depth range from the sea surface to the seabed. In order to reflect the characteristics of deep-water ocean stratification, a new parameter is introduced on the basis of the existing Vlasenko94 scheme (Formula 3), and an improved ocean stratification parameterization scheme is obtained that can better fit the ocean stratification frequency profile.

[0036] (3)

[0037] (4)

[0038] in is the floating frequency profile The last 10% of the mean value introduces the ocean stratification intensity in the deep water that is not considered in the Vlasenko94 scheme. is the maximum floating frequency, indicating the intensity of ocean stratification, and Respectively represent the layered feature depth (i.e. depth) and feature width (i.e. 0.5 According to the improved parameterization scheme, the ocean stratification parameters in the continuous system are obtained, such as Figure 2 shown.

[0039] Preferably, the real ocean stratification information can be collected not only from Argo floats but also from other field observation data, such as temperature-salinity-depth (CTD) data.

[0040] Step 2: Using the MITgcm numerical model and M4S radar imaging simulation software, different levels of sea surface wind speed are set in the radar simulation experiment to construct a sample data set of internal solitary wave vertical structure-remote sensing characteristic parameters and ocean environment parameters. The input parameters of the sample data set include five ocean stratification parameters, two internal solitary wave remote sensing characteristics and sea surface wind speed; the output includes eight vertical structure feature points. The specific process of establishing the sample data set is as follows.

[0041] The circulation model MITgcm is used to conduct a two-dimensional numerical simulation experiment of internal solitary waves in the ocean to simulate the vertical structure and horizontal flow field of internal solitary waves. The X direction of the model is the propagation direction of the internal solitary wave, and the resolution is set to 20 meters. The Z axis of the model extends vertically downward, and its length varies with the water depth. The vertical resolution is set to 10 meters, and the time step is set to 2 seconds, which meets the Courant-Friedrichs-Lewy (CFL) condition. The experiment adopts the non-hydrostatic approximation and free surface assumption, and is set to a no-slip bottom. The diffusion coefficient and viscosity coefficient are both set to 0. The initial density field of the simulation experiment is set according to the parameterized ocean stratification profile data extracted by formula (4). The initial flow field of the simulation experiment is set as a one-mode shaped internal solitary wave solution of the Korteweg-deVries (KdV) equation for a continuous stratified fluid, and according to the stratification depth characteristic parameter Set six initial amplitude value levels: 0.2 , 0.4 , 0.6 , 0.8 , , and 1.2 When the KdV solution sets the initial flow field of the model, as the model runs, the initial waveform changes, and a series of sub-waves appear to form a wave packet. The remaining stable first wave is the stable internal solitary wave solution obtained by numerical simulation. The stability judgment condition is that the amplitude of the first wave of the internal solitary wave does not change by more than 1% within 1800 time steps.

[0042] Based on the sea surface internal solitary wave flow field obtained by numerical simulation, the internal solitary wave remote sensing imaging simulation was carried out using M4S radar imaging simulation software. The radar parameters were set to C band, VV polarization, 30 o Angle of incidence, 0 o The wind speed conditions are set to 4 m / s, 6 m / s, 8 m / s, 10 m / s and 12 m / s. Based on the internal solitary wave backscattering cross-section data obtained by the M4S model, two remote sensing characteristic parameters of the internal solitary wave can be extracted through formulas (5) and (6): the distance between the brightest point and the darkest point on the satellite remote sensing image and the internal solitary wave modulation depth .

[0043] (5)

[0044] (6)

[0045] Where x is the propagation direction of the internal solitary wave, )and ) represent the minimum and maximum backscattering cross-section gradients in the x-direction. and They are the backscattering cross sections of the darkest and brightest points on the bright and dark strip radar signals of the internal solitary wave, respectively.

[0046] According to the vertical structure of internal solitary waves obtained by numerical simulation experiments of MITgcm, 8 characteristic points are extracted to replace the vertical structure as the output data of the sample data set: maximum amplitude ( ), Maximum Amplitude Depth ( ), the depth at which the amplitude is 0.25 times the maximum amplitude ( and ), the depth at which the amplitude is 0.5 times the maximum amplitude ( and ), the depth at which the amplitude is 0.75 times the maximum amplitude ( and ),like Figure 3 shown.

[0047] Preferably, the MITgcm model can be replaced by other models, such as the Regional Ocean Modeling System (ROMS) model and the global ocean circulation model LASG / IAP Climate system Ocean Model (LICOM). In addition, a sample data set of internal solitary wave vertical structure-remote sensing feature pairs can be established by collecting satellite observation-field observation data pairs.

[0048] Step 3: Use the sample data set to train the back propagation neural network to obtain the trained internal solitary wave vertical structure inversion model. Input the established internal solitary wave vertical structure-remote sensing characteristic parameters and ocean environment parameters into the sample data set such as Figure 4 In the back propagation neural network shown. The network includes an input layer, three hidden layers and an output layer. The input layer includes 8 input nodes. The number of nodes in each hidden layer is 8, and the output layer includes 8 output nodes. By training the back propagation neural network, the establishment of a nonlinear relationship between input and output is achieved. The present invention realizes the inversion of the vertical structure of the ocean internal solitary wave under a continuous layered system based on synthetic aperture radar images by combining numerical simulation and back propagation neural network. Traditional research methods are limited to a two-layer ocean model, and the vertical structure of the internal solitary wave obtained is only the amplitude at the interface between the two layers, which is very different from the real ocean environment.

[0049] The sample data set randomly assigns 75% of the data pairs as training sets and the remaining 25% as test sets. The hidden layer of the network uses the transfer function of transition, and the output layer uses the transfer function of purelin. The network training uses the trainlm training function. In addition, the learning function and performance function of the network are learngdm and mean square error MSE respectively. In order to ensure the convergence of the network and improve the convergence speed, the input and output parameters must be normalized:

[0050] (7)

[0051] in, Represents an input or output value, is the corresponding normalized value. and They are The training of the neural network will not be terminated until the MSE is less than 0.0001 or the training iteration exceeds 1000 times.

[0052] Preferably, the back propagation neural network can be replaced by other machine learning models or deep learning models. In addition, a data set is established by combining numerical simulation and radar simulation software, and a back propagation neural network is trained based on the data set to obtain an inversion model. This method can also be applied to the inversion of other internal solitary wave dynamic characteristics, such as internal solitary wave velocity and wave-induced flow field.

[0053] Step 4: Obtain remote sensing characteristic parameters of solitary waves within the target, ocean stratification parameters at the corresponding location, and sea surface wind field data.

[0054] Obtain the SAR image that records the internal solitary wave signal of the target to be inverted and perform preprocessing. The preprocessing process includes radiation correction, speckle filtering and geographic correction. Select the position of the internal solitary wave of the target to be inverted on the preprocessed image and intercept the backscattering cross section, such as Figure 5 As shown in the figure, the distance between the brightest point and the darkest point on the satellite remote sensing image is obtained. and the internal solitary wave modulation depth , the black dots indicate the location of the QB1 field observation station, and the black solid line indicates the selected backscattering cross section. According to the acquisition time of the SAR image and the location information of the solitary wave in the target, the temperature and salinity field information of the solitary wave position in the SAR observation target is obtained based on the HYCOM (Hybrid Coordinate Ocean Model) mode data, and the ocean stratification characteristic parameters are extracted through the improved stratification parameterization scheme. At the same time, based on the wind field data product with a time resolution of hours of ERA5, the sea surface wind speed data within the temporal and spatial range of the solitary wave in the SAR observation target can be extracted.

[0055] Step 5: Input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind speed data into the trained internal solitary wave vertical structure inversion model to obtain the vertical structure characteristic points of the target internal solitary wave, and finally obtain the target internal solitary wave vertical structure at full water depth through the interpolation method.

[0056] In summary, the present invention proposes an improved parameterization scheme for continuous ocean stratification to provide real ocean stratification information. For the first time, the internal solitary wave modulation depth is used as the input parameter for the inversion of the internal solitary wave vertical structure, rather than using the distance between the brightest and darkest points in the remote sensing image. As a single internal solitary wave remote sensing characteristic parameter, the internal solitary wave characteristic inversion is performed, such as Figure 6 As shown, the internal solitary wave numerical simulation is performed using the parameterized stratified profile as the initial density field, and the internal solitary wave-induced isopycnic line offset obtained is highly consistent with the field measurement results, wherein the field measurement is a solid line, and the internal solitary wave-induced isopycnic line offset obtained by simulating the parameterized buoy frequency profile based on the improved parameterization scheme is a dotted line. Based on the ocean stratification parameters and internal solitary wave remote sensing characteristic parameters provided by the improved parameterization scheme, the present invention designs a back propagation neural network model, establishes the relationship between the vertical structure characteristics of the internal solitary wave in the continuously stratified water body and the satellite image signal, and realizes the inversion of the vertical structure of the internal solitary wave in the continuously stratified system based on remote sensing images for the first time. Compared with the existing theoretical methods, the inversion model provided by the present invention is highly consistent with the field data results, such as Figure 7 As shown, the field observations are black dots and the inversion results of the vertical structure inversion model are black solid lines.

[0057] In a second aspect, the present invention provides a device for inverting continuous vertical structures of internal solitary waves in the ocean by satellite remote sensing, comprising:

[0058] A stratification parameter extraction unit, used to collect real ocean stratification data and extract ocean stratification parameters according to an improved ocean stratification parameterization scheme;

[0059] The sample data set acquisition unit is used to conduct numerical simulation and radar simulation experiments, use the extracted ocean stratification parameters to set the initial density field of the numerical simulation, and construct a sample data set of the vertical structure of the internal solitary wave - remote sensing characteristic parameters and ocean environment parameters;

[0060] A model acquisition unit, used for training a back propagation neural network using a sample data set to obtain a trained internal solitary wave vertical structure inversion model;

[0061] A target parameter acquisition unit is used to obtain remote sensing characteristic parameters of solitary waves in the target, ocean stratification parameters at the corresponding position, and sea surface wind field data;

[0062] The target structure acquisition unit is used to input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind field data into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure.

[0063] In a third aspect, the present invention provides an electronic device comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned satellite remote sensing method for inverting continuous vertical structures of solitary waves in the ocean.

[0064] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned method for inverting the continuous vertical structure of solitary waves in the ocean using satellite remote sensing.

[0065] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for inverting the continuous vertical structure of internal solitary waves in the ocean using satellite remote sensing, characterized in that: The method comprises: Step 1: Collect real ocean stratification data and extract ocean stratification parameters according to the improved ocean stratification parameterization scheme; Step 2: Conduct numerical simulation and radar simulation experiments, use the extracted ocean stratification parameters to set the initial density field of the numerical simulation, set different levels of sea surface wind speed in the radar simulation experiment, and construct a sample data set of internal solitary wave vertical structure-remote sensing characteristic parameters and ocean environment parameters; Step 3: Use the sample data set to train the back propagation neural network to obtain a trained internal solitary wave vertical structure inversion model; Step 4: Obtain remote sensing characteristic parameters of solitary waves in the target and ocean environment parameters at the corresponding position, namely, ocean stratification parameters and sea surface wind field data; Step 5: Input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind field data into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure.

2. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 1, characterized in that: The step 1 comprises: collecting the ocean true temperature profile data recorded by the submerged and floating profile detection buoy and salinity profile data , calculate the corresponding density profile , and obtain the corresponding floating frequency profile through the improved continuous stratification scheme : （1） （2） （3） in, 1028 kg / m 3 , g represents the acceleration due to gravity, It is the depth range from the sea surface to the sea bottom. is the floating frequency profile The average of the last 10% from the sea surface to the seabed, is the maximum floating frequency, indicating the intensity of ocean stratification, and They represent the layered feature depth and feature width respectively.

3. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 1, characterized in that: The step 2 comprises: Using the circulation model MITgcm, a two-dimensional ocean internal solitary wave numerical simulation experiment was conducted to simulate the vertical structure and horizontal flow field of the internal solitary wave. Based on the flow field caused by internal solitary waves on the sea surface obtained by numerical simulation experiments, the M4S radar imaging simulation software is used to simulate the remote sensing imaging of internal solitary waves. Different levels of sea surface wind speeds are set in the radar simulation experiment to extract the remote sensing image characteristic parameters of internal solitary waves. The internal solitary wave remote sensing image characteristic parameters, ocean stratification parameters, and the corresponding sea surface wind speed data set in the simulation experiment are used as the input data of the sample data set; According to the vertical structure of internal solitary waves obtained by numerical simulation experiments of MITgcm, 8 characteristic points are extracted to replace the vertical structure as the output data of the sample data set.

4. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 1, characterized in that: The step 3 comprises: The established sample data set of internal solitary wave vertical structure-remote sensing characteristic parameters and marine environment parameters is input into a back-propagation neural network, wherein the back-propagation neural network comprises an input layer, three hidden layers and an output layer, wherein the input layer comprises 8 input nodes, the number of nodes in each hidden layer is 8, the output layer comprises 8 output nodes, the hidden layer uses the transition transfer function, the output layer uses the purelin transfer function, the network is trained using the trainlm training function, the learning function and the performance function are learngdm and mean square error MSE respectively, and the training of the neural network is terminated when the mean square error MSE is less than 0.0001 or the training iterations exceed 1000 times.

5. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 1, characterized in that: The step 4 comprises: Acquire SAR images that record the solitary wave signals within the target to be inverted, and perform preprocessing procedures such as radiation correction, speckle filtering, and geographic correction; The position of the target internal solitary wave to be inverted is selected on the preprocessed image, and the backscattering cross section is intercepted to obtain the distance between the brightest point and the darkest point on the satellite remote sensing image. and the internal solitary wave modulation depth ; According to the acquisition time of SAR images and the location information of solitary waves within the target, the temperature and salinity field information of the solitary waves within the SAR observation target is obtained using the HYCOM mode data set, and the ocean stratification parameters are extracted through an improved stratification parameterization scheme; Based on the wind field data product with a time resolution of hours of ERA5, the sea surface wind speed data within the temporal and spatial range of the solitary waves in the SAR observation target are extracted.

6. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 1, characterized in that: The step 5 comprises: inputting the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind speed data into the trained internal solitary wave vertical structure inversion model to obtain the vertical structure characteristic points of the target internal solitary wave, and obtaining the target internal solitary wave vertical structure at full water depth by interpolation method.

7. The method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing according to claim 3, characterized in that: The remote sensing characteristic parameters for extracting internal solitary waves include: Extract the distance between the brightest and darkest points on satellite remote sensing images and the internal solitary wave modulation depth : （5） （6） Where x is the propagation direction of the internal solitary wave, )and ) represents the minimum and maximum backscattering cross-section gradients in the x-direction, and They are the backscattering cross sections of the darkest and brightest points on the bright and dark strip radar signals of the internal solitary wave, respectively.

8. A satellite remote sensing ocean internal solitary wave continuous vertical structure inversion device, characterized in that: include: A stratification parameter extraction unit, used to collect real ocean stratification data and extract ocean stratification parameters according to an improved ocean stratification parameterization scheme; The sample data set acquisition unit is used to conduct numerical simulation and radar simulation experiments, use the extracted ocean stratification parameters to set the initial density field of the numerical simulation, and construct a sample data set of the vertical structure of the internal solitary wave - remote sensing characteristic parameters and ocean environment parameters; A model acquisition unit, used for training a back propagation neural network using a sample data set to obtain a trained internal solitary wave vertical structure inversion model; A target parameter acquisition unit is used to obtain remote sensing characteristic parameters of solitary waves in the target, ocean stratification parameters at the corresponding position, and sea surface wind field data; The target structure acquisition unit is used to input the extracted target internal solitary wave remote sensing characteristic parameters, ocean stratification parameters and sea surface wind field data into the trained internal solitary wave vertical structure inversion model to obtain the target internal solitary wave vertical structure.

9. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; Wherein, when one or more programs are executed by the one or more processors, the one or more processors implement the method for inverting continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: Executable instructions are stored thereon, and when the instructions are executed by the processor, the processor can implement the method for inverting the continuous vertical structure of internal solitary waves in the ocean by satellite remote sensing as described in any one of claims 1-7.

Images