Analysis of Wave-Contaminated XCP Velocity Measurements

XCP velocity data, collected during severe storms to estimate wind-generated currents, require special estimation procedures. The short measurement records are dominated by large amplitude, wave-induced motions. Nonlinear likelihood estimation procedures, based on an auto-regressive description of the wave-noise, are used to estimate the wave amplitude, complex frequency and background-current model parameters. Probability distributions of the model parameters are developed from Bayesian inference methods, and used to define probability envelopes of background current profiles.