Analysis of wave directional spreading by Bayesian parameter estimation

A spatial army of wave gauges installed on an observation platform has been designed and arranged to measure the local features of winter monsoon directional waves off Taishi coast of Taiwan. A new method, named the Bayesian Parameter Estimation Method (BPEM), is developed and adopted to determine the main direction and the directional spreading parameter of directional spectra. The BPEM could be considered as a regression analysis to find the maximum joint probability of parameters, which best approximates the observed data from the Bayesian viewpoint. The result of the analysis of field wave data demonstrates the highly dependency of the characteristics of normalized directional spreading on the wave age. The Mitsuyasu type empirical formula of directional spectrum is therefore modified to be representative of monsoon wave field. Moreover, it is suggested that S_max, could be expressed as a function of wave steepness. The values of S_max decrease with increasing steepness. Finally, a local directional spreading model, which is simple to be utilized in engineering practice, is proposed.