WAVE DIRECTION SPECTRUM EXTRACTION METHOD BY SHORT- WAVE OCEAN RADAR

摘要

PROBLEM TO BE SOLVED: To develop algorithm using a premise where a nonlinear inverse problem is accurately solved, and to solve a problem regarding transcendental conditions and a weight coefficient by introducing a statistical model. SOLUTION: In a wave direction spectrum extraction method for extracting the direction spectrum of a wave from a Doppler spectrum of sea-surface scattering being observed by at least two sets of short-wave ocean radars, the direction spectrum of the wave is indicated by an index function X for developing by the coordinates of an azimuth and a frequency, and a Bayes model is introduced for extracting the direction spectrum of the wave from the Doppler spectrum. Furthermore, as the advance distribution of the Bayes model, the priori conditions of the smooth distribution of the distribution of the direction spectrum of the wave is incorporated, the weight coefficient of the priori conditions is determined (S10) by minimizing a Bayes-type information standard ABIC, and the direction spectrum is successively and approximately obtained by the maximum likelihood method of Bayes.