Study On Key Techniques Of SAR-based Rogue Wave Detection

Rogue waves, also known as extreme waves or freak waves, has extraordinary wave heights or wave shape. It is the front-line of ocean wave research in recent20years due to the catastrophic damages it caused to navigation and oil drilling. But limited by the field observations, records of rogue waves are rare and usually only time series on one single point can be obtained. Synthetic Aperture Radar (SAR) is capable of capturing high-resolution micro wave images; it is highly penetrative and can work under all weather/climate condition. Thus, it is promising to study and predict rogue wave by developing SAR-based rogue wave detection techniques.Nonlinearity is a remarkable character of rogue waves. And electromagnetic scattering feature is the key of studying SAR-based remote sensing technology. This article shows deep studies of several key technologies on SAR-based rogue wave detection with the order of:build up nonlinear ocean model and analysis the nonlinear characters between waves→rogue wave Electromagnetic backscattering (EM) model→SAR imaging simulation of rogue wave contained ocean scenes→detect rogues waves from SAR imageries→develop a software system for researchers who interested in rogue waves and those who want to detected rogue waves from SAR images. The work of this thesis includes:1. Chapter3builds up a non-linear ocean surface representative model. The basic principal of oceanic micro wave remote sensing is the adjusting of Ocean long wave to short wave. When the wave length of the ocean surface wave is longer than the resolution of radar, the adjustment between long and short waves can also determine the imaging of SAR. Although the nonlinear interaction between a single short gravity wave and a long wave can be solved analytically, the solution is less tractable in more general cases involving multiple short waves. In this work we present a numerical method of studying nonlinear interactions between a long wave and multiple short harmonic waves in infinite depth. Specifically, this method is applied to the calculation of the temporal and spatial evolution of the surface elevations in which a given long wave interacts with several short harmonic waves. Another important application of our method is to quantitatively analyse the nonlinear interactions between an arbitrary short wave train and another short wave train. From simulation results, we obtain that the mechanism for the nonlinear interactions between one short wave train and another short wave train leads to the energy focusing of the other short wave train. This mechanism occurs on wave components with a narrow frequency bandwidth, whose frequencies are near that of the later wave train.2. Doppler frequency causing by linear and nonlinear sea surface motion are deduced in chapter4. Numerical results show that Doppler frequencies distribution in SAR range direction with useful characters. Doppler frequencies causing by sea surface moving along horizontal direction of nonlinear sea surface has the same result as those of linear sea surface. It is one of basic theories of moving targets detection from sea surface. Comparing with linear sea surface, Doppler frequencies characters caused by vertical moving nonlinear sea surface can provide more characters of local sea surface patches. This finding can be used to detect targets related to sea surface or rogue waves.3. To study the electromagnetic (EM) backscatter characteristics of freak waves in moderate incidence angles, we establish an EM backscattering model for freak waves in1+1dimensions deep water. The nonlinear interactions between freak waves and Bragg short waves are considered as the basic hydrodynamic spectra modulation mechanism in the model. Numerical results suggest that the EM backscattering intensities of freak waves are less than those from the background sea surface in moderate incident angles. Compared with that of background sea surface, the normalized radar cross section (NRCS) from the rogue wave facet with the strongest slope is extremely low, and the NRCS of rogue wave area has much stronger perturbation.4. Based on the previous findings, a NRCS-intensity-perturbation-based algorithm is proposed to detect rogue waves from SAR imageries, and it is verified by processing with240SAR images. Then, a software system named' Rogue Wave Imaging Simulating and Detecting' is designed, which can be used to simulate1-D and2-D rogue waves, simulate SAR images with various radar parameters, read different SAR imageries, inverse ocean spectra from SAR imageries, and detect rogue waves from SAR imageries.