Near-shore terrain inversion method and device based on high-resolution No.3 synthetic aperture radar

Abstract

The invention provides a near-shore terrain inversion method and device based on a GF-3 synthetic aperture radar. The method comprises the following steps: acquiring original GF-3 synthetic aperture radar near-shore image data and GEBCO water depth data; inverting a sea wave spectrum of the SAR image region based on an SAR sea wave spectrum inversion algorithm; storm and surge separation is carried out through the segmentation frequency of the sea waves, and storm and surge separation is achieved; and through a near-shore terrain inversion algorithm, performing near-shore terrain inversion and verifying the accuracy of the water depth. The inversion method provided by the invention provides an effective method for SAR (Synthetic Aperture Radar) nearshore terrain inversion to solve the technical problem of SAR nearshore terrain inversion.

Description

Technical Field

[0001] The present invention relates to the technical field of synthetic aperture radar, and in particular to a method and device for inverting nearshore terrain based on a GF-3 synthetic aperture radar. Background Art

[0002] Topography is of great significance in geodetic and oceanographic research. In particular, when using numerical models for sea dynamic forecasting, topography is basically treated as seabed conditions. Traditionally, the water depth data obtained by the multi-beam echo sounder on the airborne Doppler current profiler (ADCP) can be used to construct a high-precision underwater terrain model. However, ADCP-based topography measurement requires a lot of manpower. Remote sensing can provide rich sea surface information. In addition, the ocean color of coastal waters can be inverted by remote sensing, namely water temperature, suspended sediments, and chlorophyll-a. In the field of optical remote sensing, the underwater topography inversion models in coastal areas can be roughly divided into three categories: spectral matching models, statistics-based models, and machine learning models. However, for water bodies beyond a certain depth (possibly more than 15m), it is difficult for optical remote sensing to obtain underwater topography because sunlight cannot penetrate the turbid coastal waters.

[0003] The changes in sea surface features modulated by underwater topography are captured by synthetic aperture radar (SAR). Underwater topography inversion methods based on SAR images are divided into three categories: methods based on backscattering simulation, methods based on theoretical models, and methods based on shallow water deformation. Both backscattering simulation and theoretical models simplify complex physical processes, and the inversion accuracy is poor. At present, considering the conservation of shallow sea wave periods, the commonly used method is based on shallow water deformation, which has high accuracy when the water depth is less than 30m. However, the reality is that with the rapid decrease in water depth, surges are more likely to break, and long-wave (i.e., length <600m) mapping on SAR is believed to be determined by nonlinear velocity aggregation. Summary of the invention

[0004] In view of the defects in the prior art, the purpose of the present invention is to provide a nearshore terrain inversion method and device based on GF-3 synthetic aperture radar. The present invention uses a deep learning method to invert the one-dimensional wave spectrum of GF-3 SAR images, and establishes a new nearshore water depth inversion method for SAR images, which is conducive to the use of SAR images in nearshore areas.

[0005] To solve the above problems, the technical solution of the present invention is:

[0006] A method for inverting nearshore terrain based on GF-3 synthetic aperture radar comprises the following steps:

[0007] Obtain raw GF-3 synthetic aperture radar nearshore image data and GEBCO water depth data;

[0008] Invert the wave spectrum of the SAR image area based on the SAR wave spectrum inversion algorithm;

[0009] Separation of wind, wave and swell is achieved by dividing the waves at different frequencies;

[0010] The nearshore topography inversion algorithm is used to invert the nearshore topography and verify the accuracy of the water depth.

[0011] Preferably, the step of obtaining original GF-3 synthetic aperture radar nearshore image data and GEBCO water depth data specifically includes: collecting GF-3 synthetic aperture radar data in QPSI mode between 2016 and 2023, the data location is near the Zhoushan waters, and also using water depth data provided by the open data source GEBCO, with a spatial resolution of 1 / 120° spatial resolution.

[0012] Preferably, the step of inverting the wave spectrum of the SAR image area based on the SAR wave spectrum inversion algorithm specifically includes: performing radiation calibration and geo-registration processing on the acquired GF-3 image data, and calculating the polarization spectrum and corresponding multiple modulation functions through the image, and inverting the one-dimensional wave spectrum corresponding to the GF-3SAR image through the established MLP neural network model.

[0013] Preferably, the step of separating wind waves and swell waves by dividing the waves by the frequency of sea wave segmentation to achieve separation of wind waves and swell waves specifically includes: when the waves exist in the form of mixed waves, the wave spectrum will present a double-peak or multi-peak form, each spectrum peak corresponds to the spectrum peak value of a different wave system, the trough between two adjacent peaks is considered to be the dividing boundary between the two wave systems, a wave spectrum is divided into different wave systems, each system component is judged, and separation of wind waves and swell waves is achieved.

[0014] Preferably, the step of inverting the nearshore topography and verifying the accuracy of the water depth by using the nearshore topography inversion algorithm specifically includes: based on the separated wind wave and swell wave spectrum results, substituting them into the dispersion relationship in the traditional linear waves, establishing the SAR nearshore topography inversion algorithm through an iterative process, and comparing the results with the GEBCO water depth data to verify the accuracy of the SAR terrain inversion algorithm.

[0015] Furthermore, the present invention also provides a nearshore terrain inversion device based on the GF-3 synthetic aperture radar, the device comprising a processor and a memory for storing executable instructions of the processor, the processor being configured to execute the nearshore terrain inversion method based on the GF-3 synthetic aperture radar as described above by executing the executable instructions.

[0016] Compared with the prior art, the present invention discloses a method for inverting the nearshore water depth using China's GF-3 synthetic aperture radar image, obtains the nearshore water depth from the GF-3 image, uses the established MLP neural network model to invert the one-dimensional wave spectrum corresponding to the GF-3SAR image, divides the wave spectrum derived by SAR into wind waves and swell waves, and uses the wind wave spectrum and swell wave spectrum to iteratively solve the water depth according to the wave dispersion relationship in the gentle slope process, thereby solving the problem of SAR nearshore water depth inversion and facilitating the use of SAR images in nearshore areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0018] Figure 1 The present invention is a flowchart of a method for inverting nearshore terrain based on the GF-3 synthetic aperture radar;

[0019] Figure 2 This is the water depth data map of GEBCO;

[0020] Figure 3 This is the GF-3 synthetic aperture radar image collected;

[0021] Figure 4 This is the algorithm flow chart of MLP;

[0022] Figure 5 This is the one-dimensional wave spectrum result diagram on the GF3 SAR image inverted by the MLP algorithm;

[0023] Figure 6a This is a comparison diagram of wind and wave inversion results and GEBCO water depth data;

[0024] Figure 6b This is a comparison diagram of the surge inversion results and the GEBCO water depth data. DETAILED DESCRIPTION

[0025] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0026] Specifically, the present invention provides a method for inverting nearshore terrain based on GF-3 synthetic aperture radar, such as Figure 1 As shown, the method comprises the following steps:

[0027] S1: Acquire original GF-3 synthetic aperture radar nearshore image data and GEBCO water depth data;

[0028] Specifically, the QPSI mode GF-3 synthetic aperture radar (SAR) data from 2016 to 2023 were collected near the Zhoushan waters. In addition, the water depth data provided by the open data source GEBCO was directly used, with a spatial resolution of 1 / 120°. The geographical location of the GF-3 data is as follows: Figure 2 As shown in the figure, the location of the GF-3SAR image is shown by the black border on the figure. In addition, the depth data of the General Bathymetric Chart of the Oceans (GEBCO) is directly used as the background field to show the depth of the SAR data area. Figure 3 Shows example results of observations made using GF-3SAR data over the study area.

[0029] S2: Invert the wave spectrum of the SAR image area based on the SAR wave spectrum inversion algorithm;

[0030] Specifically, all GF-3 synthetic aperture radar data are acquired, and the GF-3 image data is processed by radiometric calibration and geo-registration. The polarization spectrum and the corresponding multiple modulation functions (MTF) are calculated through the image. The one-dimensional wave spectrum corresponding to the GF-3SAR image is inverted through the established MLP neural network model. Figure 4 is a schematic diagram of the MLP model. In addition, Figure 5 This is the result of the one-dimensional wave spectrum inverted from the SAR image. It can be found that there are obvious wind waves and swell systems in the one-dimensional wave spectrum inverted from the image.

[0031] S3: Separation of wind waves and swells is achieved by dividing the frequency of sea waves;

[0032] Specifically, in this embodiment, the method proposed by Portilla et al. (2009) is used to separate wind waves from swells. The specific method is that when the waves exist in the form of mixed waves, the sea temperature spectrum will present a double-peak or multi-peak form, and each spectrum peak corresponds to the spectrum peak value of a different wave system. The "wave trough" between two adjacent peaks can be considered as the dividing boundary of the two wave systems. Through this method, a wave spectrum is divided into different wave systems, and each system component is judged to achieve the separation of wind waves and swells. Portilla et al. proposed that the parameter γ represents the ratio of the spectrum peak value of the wave system to the spectrum peak value of the PM spectrum with the same frequency as the wave system spectrum peak value:

[0033]

[0034] Among them, S(fp) is the spectral peak value of a single wave system, fp is the spectral peak frequency of a single wave system, SPM(fp) is the spectral peak value of the PM fully grown wind wave spectrum with a spectral peak frequency of fp. The basic form of the PM spectrum is as follows:

[0035]

[0036] If γ>1, the wave satisfies the wind wave overshoot phenomenon, that is, the peak is considered to be a wind wave, and if γ<1, the peak is considered to be a swell wave. Based on the above method, the one-dimensional spectrum of the inverted sea waves on the SAR image is separated into wind waves and swell waves to obtain the corresponding wind wave and swell wave results.

[0037] S4: Through the nearshore terrain inversion algorithm, the nearshore terrain is inverted and the accuracy of the water depth is verified.

[0038] Specifically, the SAR water depth data is inverted based on the wave spectrum results of the above-mentioned separated wind waves and swell waves. The specific method is: the dispersion relationship corresponding to the linear wave theory is as follows:

[0039] σ 2 =gktanhkh

[0040] Among them, σ (= 2π / T) is the circular frequency of the wave (T is the wave period); g is the acceleration of gravity; k (= 2π / L) is the wave number (L is the wavelength); h is the water depth, so the above formula can be rewritten as:

[0041]

[0042] When waves propagate from the open sea to the shore, they are affected by water depth, topography, buildings, etc., which leads to the strengthening of nonlinear effects. Therefore, the nonlinear dispersion relationship is considered:

[0043]

[0044] Where A is the amplitude, which is equal to wave height / 2. However, the water depth h cannot be expressed by a function, so an iterative method is used to approximate the true value. The specific equation is as follows:

[0045]

[0046] Through the above method, the nearshore terrain data are acquired for the wind wave spectrum and swell wave spectrum data on the SAR image respectively. Figure 6a , Figure 6b The difference between the nearshore water depth data obtained on the GF-3SAR image and the GEBCO water depth data is shown. Figure 6a is the wind and wave inversion result, Figure 6bIt is the result of surge inversion. As can be seen from the figure, the difference between the terrain depth data inverted by GF-3SAR based on wind wave spectrum and NDBC data is less than 2m, and is better than the inversion result of surge spectrum. The present invention establishes a new SAR image nearshore water depth inversion algorithm based on wind wave spectrum and linear dispersion relationship data obtained from SAR data, which is helpful for the use of SAR images in nearshore areas.

[0047] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for inverting nearshore terrain based on GF-3 synthetic aperture radar, characterized in that: The method comprises the following steps: Obtain raw GF-3 synthetic aperture radar nearshore image data and GEBCO water depth data; Invert the wave spectrum of the SAR image area based on the SAR wave spectrum inversion algorithm; Separation of wind, wave and swell is achieved by dividing the waves at different frequencies; The nearshore topography inversion algorithm is used to invert the nearshore topography and verify the accuracy of the water depth.

2. The method for inverting nearshore terrain based on GF-3 synthetic aperture radar according to claim 1 is characterized in that: The steps of obtaining the original GF-3 synthetic aperture radar nearshore image data and GEBCO water depth data specifically include: collecting GF-3 synthetic aperture radar data in QPSI mode between 2016 and 2023, the data location is near the Zhoushan waters, and also using water depth data provided by the open data source GEBCO, with a spatial resolution of 1 / 120° spatial resolution.

3. The method for inverting nearshore terrain based on GF-3 synthetic aperture radar according to claim 1 is characterized in that: The step of inverting the wave spectrum of the SAR image area based on the SAR wave spectrum inversion algorithm specifically includes: performing radiation calibration and geo-registration processing on the acquired GF-3 image data, calculating the polarization spectrum and corresponding multiple modulation functions through the image, and inverting the one-dimensional wave spectrum corresponding to the GF-3SAR image through the established MLP neural network model.

4. The method for inverting nearshore terrain based on GF-3 synthetic aperture radar according to claim 1 is characterized in that: The steps of separating wind waves and swell waves by dividing the waves by the frequency of the waves specifically include: when the waves exist in the form of mixed waves, the wave spectrum will present a double-peak or multi-peak form, each spectrum peak corresponds to the spectrum peak value of a different wave system, the trough between two adjacent peaks is considered to be the dividing boundary between the two wave systems, a wave spectrum is divided into different wave systems, each system component is judged, and the separation of wind waves and swell waves is achieved.

5. The method for inverting nearshore terrain based on GF-3 synthetic aperture radar according to claim 1, characterized in that: The steps of inverting the nearshore topography and verifying the accuracy of the water depth by using the nearshore topography inversion algorithm specifically include: based on the separated wind wave and swell wave spectrum results, substituting them into the dispersion relationship in the traditional linear waves, establishing the SAR nearshore topography inversion algorithm through an iterative process, and comparing the results with the GEBCO water depth data to verify the accuracy of the SAR topography inversion algorithm.

6. A nearshore terrain inversion device based on GF-3 synthetic aperture radar, characterized in that: The device includes a processor and a memory for storing executable instructions of the processor, and the processor is configured to execute the nearshore terrain inversion method based on the GF-3 synthetic aperture radar as described in any one of claims 1 to 5 by executing the executable instructions.

Images