Practical comparative performance survey of methods used for estimating directional wave spectra from heave-pitch-roll data

Twelve methods used for estimating the directional wave spectrum from heave-pitch-roll data are compared on realistic numerical simulations. These methods are based on different modeling approaches including simple and sophisticated ones: Fourier series decomposition. Fit to unimodal or bimodal parametric models, Variational fitting technique, Maximum Likelihood Methods, Eigenvector Methods, Maximum Entropy Methods and Bayesian approach. The comparison is performed in terms of practical aspects such as estimation error, computation time, implementation difficulty. This study must be regarded as a preliminary step devoted to the choice of optimal methods for operational in situ measurements.