Study On The Extraction Of Ocean Wave Parameters From Spaceborne Scatterometer

Ocean wave is one of the most important Ocean dynamic parameters. Up to now, the synthetic aperture radar is the only sensor to be used to measure ocean wave spectrum. However, because of the azimuthal wave cutoff and lack of power, it can not get the global high resolution ocean wave spectrum and its application is limited. The significant wave height (H1/3) can be measured globally by satellite radar altimeter. But, as its probe band of airspace is small and the two dimensional spatial resolution is too low, its applications are also limited. Therefore, it is imminent to develop an all-weather day measurement of wave parameters.In this dissertation, a method to retrieve the wave parameters from the scatterometer data will be developed. The main works are as follows, The effect of swell on radar backscattering cross sections (RCS) of the sea surface on C-band is simulated using two-scale scattering model. It with previous Stephen's works is shown that the effect of swell on the RCS decreases when radar frequency increases or radar incidence angle increases.It is shown that the relationship between the azimuth angle and the radar backscatter cross-section is a significant second-order cosine function when ocean wave parameters (significant wave height and wave steepness) and radar incidence angle are given according to buoy and scatterometer data. So we develop a back-propagation (BP) neural network algorithm to retrieve the ocean wave parameters from the scatterometer data.An algorithm to retrieve the ocean wave parameters from the ERS-1/2 and QuikSCAT scatterometer data is developed, respectively. A validation is given using NDBC Buoy data and it shows that the ocean wave parameters retrieved from scatterometer data for primary wind wave are better than those for primary swell, those for C band are better than those for Ku band and those for HH polarization are better than those for VV polarization on Ku band. Especially, wave steepness retrieved from scatterometer data is better than significant wave height and it is independent of the sea state.