PARTITIONING OCEAN WAVE SPECTRA OBTAINED FROM RADAR OBSERVATIONS

摘要

2D wave spectra of ocean waves can be partitioned intoseveral wave components to better characterize thescene. We present here two methods of componentdetection: one based on watershed algorithm and theother based on a Bayesian approach. We tested bothmethods on a set of simulated SWIM data, the Ku-bandreal aperture radar embarked on the CFOSAT (China-France Oceanography Satellite) mission which launch isplanned mid-2018. We present the results and the limitsof both approaches and show that Bayesian method canalso be applied to other kind of wave spectraobservations as those obtained with the radar KuROS,an airborne radar wave spectrometer.