The Retrieval Of Ocean Wave Information From SAR Images

The ocean wave is an important research field in Physical Oceanography. One of the key contents of the research on ocean waves is concerning ocean wave spectrum, which is closely related from ocean wave itself to the application to ocean wave theory. Unfortunately, the acquirement of two-dimensional ocean wave spectrum is too difficult to perform. Synthetic Aperture Radar (SAR) is the only instrument which is capable to acquire ocean wave information of large ocean area continuously in its orbit. So it is of scientific significance to pursue the research on how to retrieve ocean wave information from SAR images validly and correctly.It has been summarized that the imaging theory of ocean waves to SAR and the simulation and retrievement model concerning the mapping relation of ocean waves to SAR image. The effect of breaking waves on SAR imaging and the role of ocean wave spectrum parameters to mapping relation are firstly investigated. A new algorithm for retrieving ocean wave directional spectrum from SAR images, named Parameterized First-guess Spectrum model, has been proposed. And then this model has been used to the retrievement of ocean wave information from SAR images for several sea states.SAR is a microwave sensor with high resolution which is gained by numerically synthesizing the reflected signal from sea surface in its circulating route. The short wave portion of ocean wave makes the electromagnet wave to reflect through the Bragg resonance mechanism, while the long wave portion modulates short waves, making itself appeared in SAR images. According the degree of velocity bunching modulation, the imaging relationship from ocean wave spectrum to SAR spectrum can be nonlinear or linear. In this dissertation, a general nonlinear mapping relation is established and the effect of different spectral pattern and spectral parameters on simulated SAR spectrum is investigated. It is demonstrated that, in wavenumber domain, the effect of imaging nonlinearity not only causes the cutoff of high wavenumber in azimuth direction, but also affect the azimuth angle of the SAR spectrum peak to deviate the azimuth angle of the ocean wave spectrum peak through an 'extrusion ' action in azimuth direction. The degree of deviation is related to the magnitude of angel spreading coefficient s . The larger the angle coefficient of ocean wave spectrum, the more consistency between the position of the peak of ocean wave spectrum to that of the SAR spectrum corresponding to ocean wave spectrum through nonlinear mapping relation.The fundamental theory of the two prevailing models for retrieving ocean wavespectra from SAR images, Max Planck Institute model (MPI) and Semi-Parametric Retrieving Algorithm (SPRA), have been analyzed. The advantages and shortcomings of the two retrieval models have been presented and at the same time, a new improved retrieving model, the Parameterized First-guess Spectrum model, has been proposed. In this new model, we take the concept called separating wavenumber to distinguishing the wind wave and swell in wavenumber domain according to the degree of nonlinearity of the mapping relation at the specific wavenumbers. The separating wavenumber can be calculated using the additional information of wind velocity and parameters of SAR satellite. And then the SAR spectrum can be divided into SAR spectrum of wind wave and of swell according to the result of separating wavenumber, with the method of two-dimensional Gauss high-pass and low-pass filter. To the portion of SAR spectrum which is generated by wind wave, is used to search for the most suitable parameters of ocean wind wave spectrum, including propagation direction of ocean wave, the phase speed of dominating wave and the angel spreading coefficient. The swell spectrum is acquired by directly inversing the linear relation of ocean wave spectrum to SAR spectrum given the portion of SAR spectrum generated by swell. The intercomparison of MPI, SPRA, and the Parameterized First-guess Spectrum model, is performed. The result indicates that the SAR spectrum corresponding to the best wave spectrum derive from the Parameterized First-guess Spectrum model is the best fit to SAR image spectrum.In the dissertation, as another mechanism of electromagnet wave reflection except Bragg resonance in the medium incident angle, the effect of ocean wave breaking is taken consideration. The parametric method is adopted to investigate the effect of breaking waves on the backscattering section and the ocean wave-radar modulation transfer function and the effect in SAR imaging process. The outcome is as follows: the effect of ocean wave breaking is more significant to HH polarized EM wave than to VV polarized EM wave. At the incident angle between 20° to 50° , ocean wave breaking make the modulus of tilt modulation transfer function to increase and the larger the wind speed, the higher the magnitude of increase. The magnitude of the modulus of hydrodynamic modulation transfer function calculated with the effect of wave breaking is also larger than the model in which only the Bragg mechanism is taken into consideration. The amount of increase is about 30 percent to 60 percent. The effect of ocean wave breaking on tilt and hydrodynamic modulation transfer function is mainly concentrated on the waves traveling in range direction for the weak velocity bunching modulation. Once the propagation direction of the dominating ocean waves deviate range direction, the velocity bunching is rapidly increasing, causing the effect of ocean wave breaking not to be prominent any longer.Some SAR images concerning different specific sea states have been analyzed, including pure swell situation, huge wave, the refraction of ocean waves and the ocean wave field of complex sea bottom topography. In the area of complex bottom topography, the ocean wave field can be observed by SAR, which can make up the deficiency of ocean wave model. In addition, SAR is able to observe the waverefraction from obstacles and to find out the refraction coefficient beyond the range of imaging area, which is very significant to ocean engineering.