The directional Ocean Wave Spectrum (DOWS) can be determined by remote sensing through high frequency (HF) backscatter techniques using the property that HF waves are scattered from the ocean by a resonant Bragg scattering mechanism.; In order to estimate the DOWS, a narrow beam antenna is needed. If such an antenna is not available and cannot be synthesized, then a superresolution technique is needed. Superresolution techniques are the main subject of this thesis. Specifically we study the use of superresolution spectrum estimation techniques in remote sensing of the wave-height DOWS using HF backscatter radar. The techniques investigated here are: (1) MUltiple SIgnal Classification (MUSIC). (2) Maximum Entropy (ME). (3) Maximum Likelihood (ML).; Two unbiassed estimates of the directional spectrum based on the MUSIC and the ML algorithms were developed and implemented.; Experimental data were provided by the Stanford Four Frequency Radar, which was on board the ATLANTIS-II during the course of the JASIN experiment.; The estimates of the DOWS using such techniques were compared with estimates using synthetic aperture radar. The comparison showed good agreement in estimating the mean direction of energy and the half power width.; A pitch-and-roll buoy was deployed in the area probed by the radar. This buoy measures the vertical acceleration in addition to the slope in two mutually orthogonal directions, from which five independent directional moments were estimated. The moments were compared with those estimated using the superresolution techniques. Both estimates of the moments were comparable to each other. This shows that the estimated spectrum is consistent with the buoy estimate.; The advantages of using superresolution techniques are: (i) No assumptions need to be made about the shape of the spectrum. (ii) If the ship (radar) is moving in a straight line, the spectrum estimate will not be affected as long as the ship speed is less than the phase velocity of the ocean waves.
展开▼
