A Self-Contained Identification Scheme for Eddies in Drifter and Float Trajectories

Abstract

It is becoming increasingly recognized that the eddy field plays an important?possibly dominating?role for oceanic motions in many aspects (e.g., transport of properties and risk assessment in the case of extreme events). This motivates the study of individual eddy events. In the Lagrangian coordinate system, vorticity possibly associated with eddies appears in two forms: as shear vorticity between neighboring particles, and as curvature of the trajectory of a single particle. Typical field experiments in physical oceanography using surface drifters or subsurface floats do not reach data densities high enough to produce enough encounters of drifters to calculate shear vorticity between them. However, curvature in individual tracks is easily observed. This study presents a methodology that extracts segments from within a trajectory that are ?looping,? which will be interpreted as a drifter being caught in an eddy. The method makes use of autoregressive processes, a simple type of stochastic processes, which easily enables a fit to the nonperfectly shaped trajectory data usually expected from field experiments. These processes also deliver frequency and persistence of the detected eddies by a very simple calculation, which makes the methodology highly suited for automatized scanning of larger datasets.