Enhancement of satellite synthetic aperture radar (SAR) ocean images is needed for improved predictions of nonlinear internal wave effects on acoustic amplitude, phase, and coherence. Their effective use in acoustic applications can be hampered by the false negative problem of SAR images that fail to reveal internal wave presence. Our aim is a largely automatic extraction of nonlinear internal wave features such as wave amplitude, wave front width, wave front separation, and longwave correlation length. A complex discrete wavelet transform is used to denoise contrast-enhanced SAR images, and a fingerprint recognition algorithm is applied to extract edges. The image is segmented into connected components that are analyzed to obtain estimates for nonlinear internal wave parameters. To evaluate the process, estimates are compared with temperature and other data from moorings positioned approximately 100 miles east of the New Jersey coast. Extensions of the procedure may give an acoustically useful description of nonlinear internal waves in SAR images. Assimilation of parameter estimates with volume data offers a new approach to appraising internal wave effects on proagation. [Work supported by the ONR.]

Back to the main WHOI SW06 home

	© Woods Hole Oceanographic Institution                         © University of Washington
	Last updated July 2007