SPECTRAL ESTIMATION TECHNIQUES FOR MULTILOOK SAR IMAGES OF OCEAN WAVES

When deriving a directional ocean wave spectrum from a synthetic aperture radar (SAR) image of the ocean surface, spectral analysis of the image data is a key step. In this paper, we present a quantitative comparison of spectral analysis techniques for analysis or multilook SAR image data of ocean waves. Performance is rated using a "spectral SNR," defined as the ratio of the desired spectral signature to the speckle noise as it appears in the SAR image spectrum. Spectral techniques considered cover all of those now in use (including SAR processor focus adjustment) as well as new experimental methods. It is shown that the spectral-phase-shift technique, introduced in this paper, is the best approach: It maximizes the spectral SNR, it allows resolution of the 180-degrees directional ambiguity, and it applies simultaneously to all wave components present in the scene. The results are illustrated using SAR imagery of ocean waves penetrating a floating ice field.