On the nature of oceanic eddies shed by the Island of Gran Canaria

We use hydrographic and buoy data to compare the initial temperature fields and Lagrangian evolution of water parcels in two vortices generated by the southward flowing Canary Current passing around the island of Gran Canaria Island. One vortex is anticyclonic, shed in June 1998 as the result of an incident current of about 0.05 m s −1, and the second one is cyclonic, shed in June 2005 with the impinging current estimated as 0.03 m s −1. The two vortices exhibit contrasting characteristics yet display some important similarities. The isopycnals are depressed in the core of the anticyclonic vortex, at least down to a depth of 700 m, whilst they dome up in the core of the cyclonic vortex but only down to 450 m. In the top 300 m the depression/doming of the isotherms is similar for both vortices, with a maximum vertical displacement of the isotherm of about 80 m, which correspond to temperature anomalies of some 2.5 °C at a given depth. A simple method is developed to obtain the initial orbital velocity field from the temperature data, from which we estimate peak values of 0.7 and 0.5 m s −1 for the anticyclonic and cyclonic vortices, respectively. The buoys, three for the anticyclonic vortex and two for the cyclonic one, were drougued at 100 m depth, below the surface mixed layer, and their initial velocities are consistent with the above values. In both vortices, the buoys revolve either within a central core, where the rotation rate remains stable and large for several weeks, or in an outer ring, where the rotation rate is significantly smaller and displays large radial fluctuations. Within the inner core the anticyclonic vortex has significant inward radial velocity, while the cyclonic vortex has near-zero radial mean motions. The cyclonic vortex rotates more slowly than the anticyclonic, their initial periods being 4.5 and 2.5 days, respectively. A simple axisymmetric model with radial diffusion (coefficient K h≅25 m 2 s −1) and advection reproduces the observations reasonably well, the diffusive effect being more important than that resulting from the observed radial advection. The model also supports the hypothesis that the rotation rate of cyclonic vortices is less than that of anticyclonic vortices, as otherwise they would become inertially unstable. Both the buoys data and sea surface temperature images confirm that the vortices evolve from youth to maturity, as the cores shrink and the outer rings expands, and then to a decay stage, as the core rot