부산대학교 도서관

Statistical relationships between the glitter patterns of the sea surface versus the sea surface slopes for different wind velocities on the sea surface are presented. In order to obtain the statistical relationships and a suitable geometry, different heights of the detector and different lengths of the 1-D profiles of the sea surface were assumed using the OpenMP directives of CPU computing parallel. For this, a Gaussian and a non-Gaussian probability density function (pdf) for the sea surface slopes of the profile were considered, using a Rect and a Gaussian glitter function to model the geometry of the interaction of the sun on the sea profile. The curves have inflection points at different wind velocities on the sea surface for all the cases considered, where the results showed a similar behavior of the curves for both the glitter functions. Higher values for the variances of the intensities of the image using the Rect glitter function were found because the Gaussian glitter function contains less energy and a significant decrease in the values when the detector height increases. The behavior of the curves of variance depends on a large percentage of the angles of incidence. To apply a non-Gaussian pdf instead of a Gaussian pdf provides variance curves slightly smoothed but with similar values, so one of these pdfs can be used to represent both. The same occurs in the correlation function. The results present some geometry examples that can be used to optimize the experiment design.