Vasilisa Koshkina and Dmitry Krasilnikov: Evaluating the Rortex Criterion for Identifying Ocean Eddies in the Lofoten Basin

At the 61st seminar of the MIPT Center for Earth Sciences, Vasilisa Koshkina and Dmitry Krasilnikov presented an approach to automatic identification of mesoscale ocean eddies. Both speakers are doctoral students at the Institute of Oceanology, Russian Academy of Sciences: Vasilisa is supervised by Alexander V. Gavrikov, and Dmitry by Vladimir V. Zhmur. Their joint study transfers methods previously applied to atmospheric vortices to ocean data.

Mesoscale eddies transport heat, salt, mass, and geochemical tracers, yet their automatic delineation remains ambiguous. Many widely used methods operate on two-dimensional fields and define eddy centers and boundaries differently. The talk compared the Rortex criterion, which isolates the rotational component of a flow from shear, with other dynamic, geometric, and hybrid approaches used in oceanography.

The Lofoten Basin served as the test region. This energetic area of the Norwegian Sea contains the well-studied Lofoten Basin Eddy and numerous eddy structures at different scales. The authors used satellite-altimetry data and output from a regional numerical model. Different methods often agreed on large, intense structures, but produced substantially different numbers of candidates, sizes, and contours. These differences cannot be interpreted as proof that one method is superior: the study does not yet have an independent gold standard, and some algorithms require tuning for a particular dataset and resolution.

High-resolution model data made it possible to examine hourly variability in eddy contours and to move from two-dimensional maps toward a preliminary analysis of three-dimensional structure in the upper 500 m. The authors demonstrated that the volumes of identified structures can be estimated and discussed future calculations of heat and salt properties. Such estimates still require careful definition of a background state and additional validation.

The discussion focused on how an eddy definition and its boundary depend on the intended task. A boundary suitable for tracking may differ from one needed to estimate heat or salt transport. Proposed next steps include validating candidates against velocity fields and thermohaline structure, comparing them with observations, developing three-dimensional identification of the eddy axis, and analyzing eddy mergers, splits, and trajectories. The authors presented the reported results as an early stage of this work.