Figs. 1 to 3 show the anomaly in potential temperature θ , potential density σ and specific volume v at z=-2.5 m in the 1/30th-deg. OFES simulation. The anomaly is computed by removing from the initial field, the annual mean from the World Ocean Atlas (2005) dataset. Figs. 4 and 5 link these anomalies to the dynamics, the filament structure *via* the Rossby number and the submesoscale instabilities *via* the vertical velocity.

These figures show that we cannot reliably link the values of any of the anomalies to the Rossby number and the vertical velocities. The last two figures also show that although most of the large vertical velocities are associated with large values of Rossby number, the inverse is not true. You can have large Rossby number without large vertical velocity, in which case we have affair only to horizonal stirring but no mixed layer instabilities.