Desarrollo de un algoritmo híbrido (físico-estadístico) para el post-procesamiento de observaciones de espectros de oleaje obtenidos por el instrumento SAR (Synthetic Aperture Radar) del satélite ENVISAT en el océano Pacífico.

Abstract

The present document describes the developing of an algorithm for the wave spectra disambiguation, with data obtained by the synthetic aperture radar instrument of the ENVISAT on the Pacific Ocean. The main objective is obtaining an automatic post-processing method capable of enhancing spectra quality by tackling two main problems: noise presence and directional ambiguity. The methodology purposed introduces the use of wave spectral statistics from which the occurrence probability of individual wave systems can be estimated. In addition, partitioning technique is described and used to identify the wave systems mentioned before. The scheme presented consists of an iterative processing that improves the spectra quality by using multicriteria physical-statistical indicators. Solved spectra on the Pacific Ocean are compared with GlobWave database and model first guess spectra. The results show the advantages of the current algorithm presented over the image cross spectra methodology used in GlobWave database. Finally, the statistics of the whole track evidence that noise reduction is important with more than 75 % of partitions removed as noise. Also, disambiguated spectra show consistency, since the fact that 60% of the spectral energy is processed with a high-quality level.