Sarah Gille (UC San Diego, United States); Bruce Cornuelle (UC San Diego, United States); Matthew Mazloff (UC San Diego, United States)

A novel tidal analysis package (red_tide) has been developed to characterize low-amplitude non-stationary tidal energy and dominant tidal peaks in noisy, irregularly sampled, or gap-prone time series. Using Bayesian maximum a posteriori estimation assuming Gaussian linear statistics, we recover tidal information by expanding conventional harmonic analysis to include prior information and assumptions about the statistics of a process, such as the assumption of a red spectral background. We utilize a hierarchy of test cases, including synthetic data and observations, to evaluate this method and its relevance to analysis of model output and observational data with a tidal component and an energetic non-tidal background. Analysis of coastal radar data follows, revealing characteristics of stationary and non-stationary tides in surface current observations.